You are on page 1of 11

Physics and Chemistry of the Earth 36 (2011) 918928

Contents lists available at SciVerse ScienceDirect

Physics and Chemistry of the Earth

journal homepage:

Fish as bioindicators in aquatic environmental pollution assessment: A case study

in Lake Victoria wetlands, Uganda
I. Naigaga a,, H. Kaiser b, W.J. Muller c, L. Ojok d, D. Mbabazi e, G. Magezi e, E. Muhumuza e
Department of Wildlife and Animal Resources Management (WARM), School of Veterinary Medicine, Makerere University, P.O. Box 7062, Kampala, Uganda
Department of Ichthyology and Fisheries Science, Rhodes University, P.O. Box 94, Grahamstown 6140, South Africa
Institute for Water Research, Rhodes University, P.O. Box 94, Grahamstown 6140, South Africa
Department of Veterinary Pathology, School of Veterinary Medicine, Makerere University, P.O. Box 7062, Kampala, Uganda
National Fisheries Resources Research Institute (NaFIRRI), P.O. Box 343, Jinja, Uganda

a r t i c l e i n f o a b s t r a c t

Article history: Growing human population and industrialization have led to the pollution of most aquatic ecosystems
Available online 16 August 2011 and consequent deterioration in environmental water quality. Indicator organisms are needed to improve
assessment programmes on the ecological impacts of anthropogenic activities on the aquatic environ-
Keywords: ment. Fish have been widely documented as useful indicators of environmental water quality because
Bioindicators of their differential sensitivity to pollution. This study investigated the environmental water quality of
Fish selected wetland ecosystems using fish as biological indicators. Fish community structure in relation
Lake Victoria
to water quality was assessed in five wetlands along the shoreline of Lake Victoria from August 2006
to June 2008. Four urban wetlands were variedly impacted by anthropogenic activities while one rural
wetland was less impacted, and served as a reference site. Fish species diversity, abundance and richness
were assessed, and canonical correspondence analysis (CCA) was used to evaluate the relationship
between the fish communities and environmental variables. Results revealed that urban effluent
impacted negatively on water quality and consequently the fish community structure. A total of 29 fish
species were recorded throughout the study with the lowest number of 15 species recorded in the most
impacted site. Shannon diversity and Margalef species richness indices were highest at the references site
and lowest at the most impacted site. Wetland haplochromis species dominated the reference site, while
oreochromis species dominated the most impacted site. The inshore locations registered higher species
diversity and low species richness than the offshore locations. Low dissolved oxygen, pH, secchi depth
and high electrical conductivity, total phosphorous, and total nitrogen were strongly associated with
the effluent-impacted sites and greatly influenced the fish community structure. This study recommends
the use of fish as valuable biological indicators in aquatic environmental pollution assessment.
2011 Elsevier Ltd. All rights reserved.

1. Introduction that these heavily vegetated wetlands serve as structural and

low-oxygen refugia for fishes that can tolerate such conditions.
The Lake Victoria wetlands on the lakes fringes have a funda- They also function as barriers to the dispersal of Nile perch, thereby
mental socio-ecological function sustaining the livelihoods of rural protecting the cichlids from Nile perch predation. However, the
and suburban communities while contributing to shore stabiliza- wetland socio-ecological functions and the continued survival of
tion, flood control, water storage, purification and recharge, nutri- the fish species in such habitats are threatened by human impacts
ent retention and micro-climate regulation (Bakema and Iyango, in the catchments.
2000; NWDR, 2006; Crona et al., 2009; Swallow et al., 2009). The Due to the increasing human population size, wetlands on the
lakes wetlands and coastal shallow bays also play a role in the con- Ugandan side of Lake Victoria are being extensively degraded by
servation of fish biodiversity (Chapman et al., 2002, 2003; Balirwa human activities such as habitat modification for agriculture,
et al., 2003). Particularly, rocky and shallow inshore habitats with urbanization and infrastructure provision, municipal and animal
fringing macrophyte cover are important refugia for many endan- waste disposal, small-scale economic production enterprises such
gered indigenous fishes of Lake Victoria (Chapman et al., 2002; as informal slaughter houses, poultry units, motor garages and
Balirwa et al., 2003, 2004b; Namulemo, 2004). The authors argue car washing facilities (Odada et al., 2004; Mugisha et al., 2007;
Banadda et al., 2009; Munabi et al., 2009). These authors argue that
Corresponding author. these activities serve as point and non-point sources of pollution
E-mail address: (I. Naigaga). and lead to increased wastewater discharges that carry potential

1474-7065/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928 919

pollutants and cause deterioration of the aquatic environment. and Chapman, 2000). This makes Lake Victoria and its catchment
This is manifested through alteration of physical and chemical an important study area for fish community observation. Knowl-
variables beyond safe levels, consequently impacting on biodiver- edge on water quality indicator organisms is critical to the conser-
sity especially that of fishes (Balirwa et al., 2003; Nzomo, 2005; vation of the cichlid fauna and the regulation of biodiversity loss in
Carr and Neary, 2006). the lake. However, the relationship between water quality and fish
There has been a decline of fish stocks and fish species diversity assemblages has been poorly documented in Lake Victoria, despite
in Lake Victoria (Balirwa et al., 2003; Ogutu-Ohwayo and Balirwa, being important for water management and ecological monitoring.
2004), due to overexploitation, use of destructive fishing gear and Therefore, this study assessed urban wetland fish biodiversity that
methods, environmental degradation, and ineffective or absent is threatened by pollution from industrial and domestic effluent.
management. The decline has also been attributed to the introduc- This research investigated the fish community structure in relation
tion of the non-indigenous Nile perch. The increase in Nile perch to the water quality, with the aim of exploring the use of fish as
population size coincided with the loss of many populations of biological indicators of water quality deterioration in selected wet-
indigenous species (Balirwa et al., 2003, 2004a; Ogutu-Ohwayo lands in Uganda.
and Balirwa, 2004; Witte et al., 2007). Long-term recommenda-
tions for management of this complex socioecological challenge in-
clude putting in place effective research and management 2. Materials and methods
programmes, strong policy and legal frameworks and information
systems for management of the fisheries and fish habitats (Balirwa 2.1. Study area
et al., 2003; Machiwa, 2003; Ogutu-Ohwayo and Balirwa, 2004).
Immediate interventions included control of fishing effort and Data were collected in the bays of five wetland ecotones located
law enforcement on the use of illegal fishing gear and methods in one ecoregion along the Ugandan shoreline of Lake Victoria.
(Ogutu-Ohwayo and Balirwa, 2004) alongside re-stocking pro- These wetland ecotones included Nakivubo, Kirinya, Masese, Win-
grammes and aquaculture which is still on a small scale (FAO, day Bay and Lwanika. The Nakivubo wetland has become severely
2006). The environmental changes due to industrial and domestic degraded over recent years and is threatened by industrial activity
wastewater effluent may reduce fish access to the wetland refugia and residential development (COWI and VKI, 1998; Emerton et al.,
in urban waters. Assessment and mitigation of the environmental 1998). The wetland receives secondary treated effluent from the
impacts on water quality and biodiversity have now become Bugolobi sewerage treatment works and heavily polluted waste-
necessary. water comprising of runoff and effluent from guesthouses, slum
Efforts to prevent or reverse aquatic environmental impacts de- dwellings and industries in Kampala city through the Nakivubo
mand efficient and reliable mechanisms of detection, especially of channel, and discharges into Lake Victoria at Murchison Bay
system contamination. Thus, there is a need to establish a biolog- (Fig. 1). The Nakivubo channel is a 12.3 km long and traverses
ical monitoring system (Balirwa et al., 2003; Nzomo, 2005). Two highly populated Kampala slums, markets and industrial areas.
biomonitoring studies of the Ugandan part of Lake Victoria sup- Point source pollution and landuse-based nonpoint source pollu-
ported and recommended biomonitoring as a useful tool for mon- tion in the channel have an impact on the water quality and the
itoring water quality as it integrates responses to combinations of aquatic life in the inner Murchison Bay (Kansiime et al., 2007; Kay-
all contaminants thereby describing overall effects on the environ- ima et al., 2008; Banadda et al., 2009) where sampling for this
ment and biodiversity in a water body (Sekiranda, 2006; Birungi study was carried out. The bay receives raw sewage, municipal
et al., 2007). Birungi et al. (2007) monitored selective trace heavy and industrial wastewater discharge from Kampala (Matagi,
metals using Nile tilapia (Oreochromis niloticus) and Sekiranda 2002; Kansiime et al., 2007; Kayima et al., 2008; Banadda et al.,
(2006) employed GIS techniques to assess water quality and inver- 2009). Kirinya wetland is located along the Northern shoreline of
tebrate and fish community structure in three shallow bays. How- Lake Victoria in Jinja municipality, lying to the north east of the
ever, Lake Victoria has diverse ecological zones and the impacts of Napoleon gulf (Fig. 1). This wetland receives discharges of domes-
environmental degradation on water quality are complex and their tic, municipal and industrial effluent, surface runoff and storm
extent is poorly understood. water from Jinja town (Kelderman et al., 2007; Oguttu et al.,
Fish have been widely documented as indicators of water 2008). Although this wetland is not an official waste dumping site,
quality because of their sensitivity to pollution (Mora et al., 2003, open waste dumping of both biodegradable and non-biodegrad-
2008; Gratwicke, 2004; Gratwicke and Speight, 2005; Mora and able materials was frequently observed during this study. Kirinya
Robertson, 2005; Sekiranda, 2006; Das and Chakrabarty, 2007). wetland is also home to hotels, industries, factories, metallurgy
Estimating the number of species from a particular area remains and Jinja municipalitys wastewater treatment oxidation ponds,
a fundamental theme of ecology. Species diversity is related to the Kirinya National Water and Sewerage Corporation Oxidation
the functioning of ecological systems and helps understand the ponds that discharge municipal secondary treated effluent into
mechanisms and effects of environmental disturbances such as Lake Victoria.
pollution (McGill et al., 2007; Tokeshi and Arakaki, 2007; Mora Masese wetland is located 2 km from Kirinya wetland (Fig. 1).
et al., 2008). Estimation of species diversity is also useful for This wetland is used for open dumping of waste material in Jinja
detecting trends, impacts, or recovery of ecosystems, and quantifi- municipality, both biodegradable and non-biodegradable including
cation of extinction risks and thus prioritization of conservation of chemical waste. Masese wetland was gazetted for waste disposal
biodiversity hot-spot areas (Mora et al., 2008). Ideally, complete over three decades ago, when the area was relatively inhabited
inventories provide a full account of the species richness and diver- (Jinja District Profile, 2003). Masese wetland receives urban efflu-
sity but these are rarely available. In their place, sampling of biota ent and storm water runoff through a 1.4 km long drainage chan-
or surveys provide important data. nel. This channel carries domestic effluent from several
Fishes are ideal for ecological investigations. They are readily residential villages and informal settlements in the outskirts of Jin-
identified and are species-rich. Additionally, warm tropical waters ja Municipality and pours into the Masese wetland before empty-
with high visibility are conducive to observations (Gratwicke, ing into Lake Victoria. Winday Bay is almost adjacent on the south-
2004). The species flock of endemic haplochromine cichlids in west of the Napoleon gulf (Fig. 1). The catchment of Winday Bay is
the basin of East Africa is well known as the most diverse and less disturbed than that of Kirinya and Masese. The main activity is
extensive radiation in the evolution of vertebrates (Rosenberger crop cultivation, with recent aquaculture. Lwanika wetland, lo-
920 I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928

Fig. 1. Map showing the location of the study area. Inset is map of Uganda showing the study area.

cated within the same ecoregion, is presumed to be a rural less pol- od while NO3N was determined using the cadmium method. The
luted wetland located in Mayuge district and without any effluent water sample was passed through a column packed with cadmium
inlet (Fig. 1). For this reason Lwanika wetland was hypothesized to granules. Then the sample colour was developed with a combined
be less impacted by urban effluent than the other wetlands and sulfanilamide and 1-naphthyl-ethylenediamide dihydrochloride
was thus used as a reference site. This wetland buffers Lake Victo- and absorbance was read at 543 nm within 2 h. Total dissolved
rias Macdonald Bay against surface runoff from the catchment. nitrogen and total phosphorus concentrations were determined
following oxidation of 20 ml samples in alkaline and acidic persul-
fate, respectively.
2.2. Sampling

The study involved collecting data every 2 months from August 2.3. Statistical analysis
2006 to June 2008 over 12 field sampling trips for water quality
physical and chemical parameters and fish. Sampling was done Means, standard deviations and ranges for each physicochemi-
at two locations, i.e. the inshore location (closest to the shoreline) cal variable were calculated. Analysis of variance (ANOVA) was
and the offshore location approximately 60 m into the bay. In Mur- used to compare differences between mean water quality values
chison Bay the inshore location was close to the Nakivubo channel for different sites on Log 10 (x + 1)-transformed data. Tukeys
discharge point. In Kirinya and Masese, the inshore locations were post-hoc HSD test was used to identify significant differences be-
close to the national wastewater, and the drainage channel dis- tween means of different sites (p 6 0.05). Species richness was
charge points, respectively. There was no effluent discharge associ- estimated using the Margalef Index, as it has discriminating ability
ated with Winday Bay and Lwanika inshore locations. At each to measure the number of species present for a given number of
location, a 1-day experiential fishing was conducted. Fish were ob- individuals. Species diversity was estimated using the Shannon
tained by the use of multiple size monofilament gill nets set for 6 h Diversity Index and indices between sampling sites were com-
(4:0010:00 am), identified to species level and counted. Water pared using Randomization Test (Seaby and Henderson, 2000,
samples were collected using a Van Dorn water sampler and phys- 2006). Simpson dominance index was used to calculate species
icochemical variables were measured between 2:00 pm and abundance distributions. The Simpsons species dominance index
4:00 pm. The variables were those regularly monitored by the describes the probability that a second individual drawn from a
water resource management. They were: pH, electrical conductiv- population should be of the same species as the first (Seaby and
ity, temperature, dissolved oxygen, secchi disk depth, total phos- Henderson, 2006). The index is reported as one of the most mean-
phorus (TP), total nitrogen (TN), nitritenitrogen (NO2N), ingful and most robust diversity measures available as it captures
nitratenitrogen (NO3N), ammoniumnitrogen (NH4N), and the variance of the species abundance distribution (Magurran,
chlorophyll a (Chl-a). 2004). Cluster Analysis (CA) and Principal Component Analysis
The temperature (C), pH, dissolved oxygen (mg/l) and electrical (PCA) were used to group the sampling locations into clusters on
conductivity (lS/cm) were measured in situ using portable meters. the basis of their similarity in fish species characteristics (Manly,
Water transparency was measured using a 25 cm black and white 2005; Bierman et al., 2011).
secchi disk. TN, TP, NH4N, NO2N, NO3N, and Chl-a were mea- Hierarchical agglomerative CA by means of the Wards method
sured using a spectrophotometer following methods of Stainton using Euclidean distances as a measure of similarity was per-
et al. (1977). NH4N was measured by the indophenol-blue meth- formed on log-transformed normal data sets (Henderson and
I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928 921

Seaby, 1999, 2007). The transformation helped to standardize the

Mean standard deviation, minimum and maximum in parenthesis, total depth, secchi depth, temperature, dissolved oxygen and electrical conductivity, pH range, dissolved inorganic nitrogen, total nitrogen and total phosphorous in

4.007 2.275c
0.045 0.003c

0.060 0.018c
(0.765, 6.108)
(0.031, 0.059)

(0.031, 0.088)
data into smaller units than using species numbers which are often

128.0 7.5b
28.4 0.6a

(119, 138)
widely varied. canonical correspondence analysis (CCA) was per-
(1.95, 2.5)

(0.7, 0.98)

(26.9, 29)
2.2 0.2b

6.7 0.9b
0.9 0.1b


(5.55, 8)
formed on normal abundance data to evaluate the relationship be-
tween fish communities and environmental variables using the
Environmental Community Analysis Package 2 (ECOM) by Hender-
son and Seaby (2000). A Monte Carlo test with 1000 randomiza-

5.409 1.691c

0.063 0.013c
0.051 0.003c

(2.205, 6.945)
(0.035, 0.061)

(0.043, 0.082)
tions was run to test for the probability of the observed patterns
Lwanika Bay

133.4 5.2b
(26.5, 28.6)
27.5 0.6a

(127, 140)
(5.7, 7.53)
(0.5, 0.87)

being due to chance. Rare species, i.e., those that occurred less than
6.9 0.6b
0.8 0.1b
1.4 0.2

(1, 1.8)

four times at a site were reported in the results but omitted from
the analysis. Omission of rare species from the analysis is recom-
mended as it characterizes the main body of the community rather
than giving much weight to species that are almost never encoun-

(2.464, 60.457)
0.808 0.574d
29.35 15.83a
0.432 0.118a
473.5 145.7c

tered. This minimizes bias and reduces the error of estimation

(0.080, 1.538)
(0.024, 1.59)
(24.1, 26.2)
(0.33, 0.81)
25.5 0.7b

(Arimoro, 2009; ter Braak and Smilauer, 2002; ter Braak, 1986).
(227, 635)
(3.02, 7.9)

0.6 0.2a

5.2 1.8c
(0.9, 3.1)

2.4 0.8

3. Results

3.1. Physical and chemical variables in the study sites

(6.467, 67.531)
Murchison Bay

0.869 0.165d

0.822 0.461d
40.35 21.11a
580.2 119.5c

(0.269, 1.614)
(0.096, 2.008)
(24.1, 26.1)
(0.18, 0.49)
25.3 0.6b

(383, 707)
(1.1, 5.36)
2.4 1.4a

0.4 0.1a

Significantly lower water average temperatures (25 0.6

(0.7, 2.7)
1.4 0.5


26.8 0.8 C), dissolved oxygen concentrations (2.4 1.4

5.2 1.8 mg/l) and secchi depth readings (0.4 0.10.6 0.20 m)
were by Murchison Bay inshore and offshore sampling sites, and
Kirinya inshore than at the reference site locations (Lwanika in-
11.31 20.16b

0.082 0.079c
0.045 0.007c

(0.720, 71.22)

(0.016, 0.316)
(0.015, 0.096)
110.7 13.4b

shore and offshore) and the other sites where these variables aver-
(25.9, 27.8)
(0.54, 1.45)
27.2 0.6a

(4.9, 8.66)
7.6 1.0b
1.0 0.3b

(76, 131)

aged 27.1 0.728.4 0.6 C; 6.0 0.67.6 1.0 mg/l; and


(2.25, 5)
3.9 0.8


0.8 0.11.1 0.2 m respectively (Table 1). The Murchison Bay

sites had, a lower pH range, i.e., 4.967.2 than the other sites with
pH values of 69. The inshore and offshore locations in this bay had
a significantly higher electrical conductivity of 580.2 119.5 and
0.109 0.085bc
12.33 12.32b
0.067 0.007c

(0.913, 30.84)

(0.038, 0.311)
(0.023, 0.100)
117.8 14.5b

473.5 145.7 lS/cm, dissolved inorganic nitrogen (DIN) of

Winday Bay

(25.5, 27.8)
(0.43, 1.73)

27.1 0.7a
(0.3, 1.09)

7.4 1.0b

0.869 165 and 0.432 118 mg/l, total phosphorus (TP) of

0.8 0.3b

(79, 135)
(4.6, 8.3)
1.2 0.5

0.822 0.461 and 0.808 0574 mg/l, and total nitrogen (TN) of

40.35 21.11 and 29.35 15.23 mg/l, followed by Kirinya inshore

301.9 96.4 lS/cm, 0.374 0.133 mg/l DIN, 0.408 0.261 mg/l TP
and 34.08 35.13 mg/l TN. The reference site inshore and offshore
0.134 0.160b
0.079 0.008c
(0.025, 0.114)

(0.026, 0.606)
11.83 9.74b
120.4 18.5b

(0.81, 33.14)

locations had the lowest DIN concentration (0.051 0.003 and

(26.1, 28.9)
(0.64, 1.41)

(5.44, 9.02)

(75, 158.8)
27.3 0.8a
(1.6, 3.22)

7.2 1.1b

1.1 0.2b

0.45 0.003 mg/l), TP (0.063 0.013 and 0.060 0.018 mg/l) and
2.4 0.5


TN (5.409 1.691 and 4.007 2.275 mg/l) concentrations followed

by Winday Bay offshore and inshore locations (DIN: 0.045 0.007
and 0.067 0.007 mg/l, TP: of 0.082 0.079 and 0.109 0.085 mg/l
and TN: of 11.31 20.16 and 12.33 12.32 mg/l) (Table 1).
0.171 0.129b
32.05 22.65a
0.091 0.009c
(0.028, 0.141)

(0.028, 0.449)
136.4 24.4b

(8.02, 69.89)
the study sites. Different superscripts in a raw denote significant differences.

(0.45, 1.21)

(81, 176.7)
27.3 0.7a

(4.52, 6.7)
(0.72, 1.5)

(26, 28.6)
0.9 0.3b

6.0 0.6c
1.2 0.3

3.2. Composition and relative abundance of fish species in the study



A total of 29 fish species belonging to five families were

11.12 10.51b

0.148 0.114b
0.138 0.047b
(0.014, 0.567)

recorded, with the majority of the species (83%) belonging to the

(0.020, 0.379)
134.3 35.0b
(82.7, 192.7)

(1.305, 4.07)
(26.3, 28.6)

(5.23, 8.55)
27.6 0.7a

Cichlidae. The other four families were Alestiidae, Centropomidae,

(0.75, 1.2)

6.9 1.0b
0.9 0.1b
2.5 0.6

Protopteridae and Mochokidae (Table 3). No significant seasonal

(2, 4)

variation in fish assemblages was observed. Before excluding the

rare occurrences, Winday Bay had the highest number of species
(22) followed by Lwanika, the reference site and Masese with 20
34.08 35.13a
0.374 0.133a

0.408 0.261a
(0.027, 1.612)

(0.049, 0.766)

each (Tables 2 and 3). After exclusion of rare occurrences, reference

301.9 96.4a

(3.509, 5.22)
(25.6, 27.9)
(0.35, 0.78)
26.8 0.8a

(159, 448)
(1.26, 4.8)
(0.35, 1.6)

site had the highest number of species (12) followed by Winday

3.1 1.1a

0.5 0.1a
1.0 0.4

Bay with 11 species (Tables 2 and 4). The inshore sampling location

registered the highest number of species before excluding rare

occurrences (Tables 2 and 3). After excluding the rare occurrences,
Kirinya, Winday Bay and Murchison Bay had equal numbers of
Cond. (lS/cm)

species while the reference site and Masese had more species in
T-Depth (m)

S-Depth (m)

DIN (mg/l)
Temp. (C)

DO (mg/l)

the inshore locations (Tables 2 and 4).

TN (mg/l)

TP (mg/l)

Thirteen common occurrence species were recorded for the 10

Table 1


sampling locations and their abundance distributions calculated

using Simpsons species dominance index (Table 4). Five species
922 I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928

Table 2
Number of fish species observed in the five wetlands before and after excluding the rare occurrence species.

Site Including rare occurrences Excluding rare occurrences

Number of species Number of fish Number of species
Total Inshore Offshore Total Inshore Offshore Total Inshore Offshore
Kirinya 19 18 15 1028 683 345 10 10 10
Masese 20 16 12 996 737 259 9 9 8
Winday Bay 22 18 17 2115 1467 648 11 11 11
Murchison Bay 15 13 11 1076 487 589 7 7 7
Lwanika 20 18 12 1611 805 806 12 12 10

Table 3
Relative abundance (percentages) of fish species from different genera and families in the study wetlands. Sampling locations L In and L Off = Lwanika Inshore and Offshore
(Reference sites), M In and M Off = Masese Inshore and Offshore, W In and W Off = Winday Bay Inshore and Offshore, K In and K Off = Kirinya Inshore and Offshore, MB In and MB
Off = Murchison Bay Inshore and Offshore. N for L-In = 805, L-Off = 806, M-In = 737, M-Off = 259, W-In = 1467, W-Off = 648, K-In = 683, K-Off = 345, MB-In = 487, and MB-
Off = 589.

Family Genus Species K-In K-Off M-In M-Off W-In W-Off MB-In MB-Off L-In L-Off
Cichlidae Oreochromis O. leucostictus 8.20 2.03 12.62 0.00 3.27 0.00 35.73 30.39 0.87 0.00
O. niloticus 13.32 42.32 6.38 25.48 9.20 0.27 30.18 30.05 17.43 26.30
O. variabilis 0.00 0.00 0.00 0.39 25.02 0.09 0.00 0.00 0.00 0.00
Paralabidochromis P. red tail 23.43 9.28 14.93 5.79 33.81 0.16 6.37 13.92 28.43 16.25
P. crassilabris 5.56 1.45 6.38 1.93 2.79 0.06 0.41 0.34 7.66 15.01
P. big eye 0.00 0.00 0.00 0.39 2.18 0.04 2.67 2.55 0.00 0.00
P. chilotes 0.15 0.29 0.00 0.00 0.07 0.00 0.00 0.00 0.00 0.00
P. rock kribensis 0.00 0.00 0.00 0.00 0.07 0.00 0.00 0.00 0.00 0.00
P. long shaft 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.62 0.50
Harpagochromis H. victorianus 4.54 0.00 0.00 0.00 0.68 0.01 0.00 0.00 3.09 6.70
H. serranus 0.29 0.58 0.00 3.47 0.00 0.00 0.21 0.00 0.37 0.00
H. quiarti 0.59 0.00 1.49 0.00 0.00 0.00 0.00 0.34 0.00 0.00
Xystichromis X. phytophagus 0.73 0.00 0.95 0.00 0.00 0.01 0.00 0.00 1.48 2.73
Yssichromis Y. fusiformis 0.00 0.00 1.63 0.00 0.00 0.00 0.00 0.00 0.25 0.00
Astatotilapia A. nubila 0.15 0.58 0.14 0.00 0.07 0.00 0.00 0.00 0.37 0.00
A. latifasciata 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00
A. brownie 4.69 0.87 0.27 0.00 2.59 0.05 5.34 3.90 8.65 4.09
Astatoreochromis A. alluaudi 5.27 1.45 26.19 0.77 3.00 0.00 0.00 0.00 5.19 0.00
Lipochromis L. melanopterus 0.00 0.00 0.27 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Ptyochromis Ptyochromis sp 0.15 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Clariidae Clarias C. gariepinus 0.00 1.16 0.00 0.00 0.14 0.00 0.41 0.00 0.00 0.00

Centropomidae Lates L. niloticus 11.42 26.67 5.29 28.57 4.98 0.19 7.19 6.28 4.45 8.44

Alestiidae Brycinus B. sadleri 1.02 0.29 5.02 2.32 1.91 0.01 9.65 6.11 2.47 8.31
B. jacksonii 0.00 0.00 0.00 0.77 0.07 0.00 0.00 0.00 0.25 0.00

Mochokidae Synodontis S. afrofisheri 0.00 0.00 0.00 0.00 0.00 0.00 0.41 0.00 0.87 0.00

Protopteridae Protopterus P. aethiopicus 0.29 0.29 0.54 0.00 0.00 0.00 0.21 0.00 0.00 0.50

occurred in all sampling locations and these included O. niloticus, served only in Winday Bay. Ptyochromis sp. were only found in Kiri-
Lates niloticus, Tilapia zillii, Brycinus sadleri and Paralabidochromis nya, Lipochromis melanopterus in Masese and Paralabidochromis
redtail (Table 4). Three of these species, i.e., O. niloticus, L. niloti- long shaft at the reference site, Lwanika. Except for O. variabilis,
cus, and T. zillii had been introduced into Lake Victoria while B. sad- most of these species occurred once and were recorded in small
leri and P. redtail are endemic. The highly adaptive and resilient numbers (Table 3).
O. niloticus was the most dominant species in 7 of the 10 sampling
locations. At Lwanika offshore and Kirinya offshore P. redtail 3.3. Spatial diversity and clustering of sites based on fish species
dominated and Winday Bay offshore was dominated by L. niloticus. characteristics
The effluent-impacted Murchison Bay supported fewer fish species
and all these species were recorded in the other study sites. A num- Spatial diversity and clustering was based on common-occur-
ber of native Haplochromine species that had been observed in the rence species. Shannon diversity and Margalef species richness
less impacted sites were not observed in Murchison Bay. These indices varied significantly among sampling locations (p 6 0.05).
species were two Astatotilapia species, i.e. Astatotilapia nibula The two indices were highest at the reference site and lowest at
and Astatotilapia latifasciata, Paralabidochromis chilotes, Harpag- Murchison Bay for the inshore locations (Fig. 2). For the offshore
ochromis victorianus, Astatoreochromis alluaudi, Xystichromis phy- locations, the Shannon diversity index was highest at the reference
tophagus, and Yssichromis fusiformis. There were rare occurrences site and lowest at Masese, while the Margalef richness index was
of Oreochromis variabilis and Paralabidochromis rock kribensis ob- highest at Winday Bay and lowest at Murchison Bay (Fig. 2). With
I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928 923

Table 4
Ranks of species based on variance of the species abundance distributions calculated using Simpsons species dominance index for the common occurrence species.

Rank Lwanika (reference site) Winday Bay Murchison Bay Masese Kirinya
Inshore Offshore Inshore Offshore Inshore Offshore Inshore Offshore Inshore Offshore
1 O. niloticus P. redtail O. niloticus L. niloticus O. niloticus O. niloticus O. niloticus O. niloticus O. niloticus P. redtail
(6.409) (6.705) (8.344) (7.031) (5.230) (6.817) (10.295) (8.412) (9.202) (12.098)
2 A. brownie (5.182) L. niloticus (5.979) T. zillii O. niloticus P. redtail T. zillii T. zillii L. niloticus L. niloticus O. niloticus
(7.604) (6.542) (2.704) (6.359) (9.063) (7.127) (7.461) (9.756)
3 O. leuco. (4.500) O. niloticus (5.818) P. redtail P. redtail O. leuco. O. leuco. L. niloticus T. zillii O. leuco. P. crass.
(6.219) (6.339) (2.458) (4.562) (7.639) (6.013) (6.871) (5.000)
4 P. riponian. B. sadleri L. niloticus P. riponian. T. zillii P. redtail O. leuco. P. crass. A. alluaudi L. niloticus
(3.930) (4.013) (5.814) (4.309) (1.977) (3.992) (6.404) (5.000) (3.580) (4.521)
5 P. crass. (3.836) T. zillii P. crass. O. variabilis B. sadleri B. sadleri A. alluaudi B. sadleri B. sadleri A. alluaudi
(3.750) (4.100) (3.072) (1.969) (3.366) (5.397) (5.000) (3.500) (3.333)
6 L. niloticus X. phyto. A. alluaudi H. Victorian. A. brownie L. niloticus P. redtail P. redtail T. zillii T. zillii
(3.539) (3.726) (4.026) (2.500) (1.890) (3.073) (5.319) (4.200) (3.443) (2.677)
7 B. sadleri P. crass. P. riponian. A. brownie L. niloticus A. brownie P. crass. A. alluaudi P. riponian. P. riponian.
(3.333) (3.625) (3.178) (2.200) (1.500) (2.091) (5.051) (2.000) (2.819) (2.377)
8 H. Victorian. P. riponian. O. variabilis T. zillii P. riponian. P. riponian. A. brownie O. leuco.
(3.261) (2.708) (3.128) (1.964) (2.403) (1.000) (2.351) (2.333)
9 T. zillii H. victorian B. sadleri P. crass. B. sadleri P. redtail A. brownie
(3.067) (2.524) (2.291) (1.794) (1.521) (2.202) (1.000)
10 P. redtail (2.264) A. brownie (1.970) H. Victorian. A. alluaudi P. crass. B. sadleri
(1.607) (1.000) (1.980) (1.000)
11 X. phyto. A. brownie B. sadleri
(2.200) (1.400) (1.000)
12 A. alluaudi

Fig. 2. Shannon diversity H and Margalef richness D indices for fish species in the ten sampling locations. Where: L In = Lwanika Inshore (reference site), W In = Winday Bay
Inshore, M In = Masese Inshore, MB In = Murchison Bay Inshore and K In = Kirinya Inshore. L Off = Lwanika Offshore (reference site), W Off = Winday Bay Offshore, M
Off = Masese Offshore, MB Off = Murchison Bay Offshore and K Off = Kirinya Offshore.

the exception of the reference site, the offshore sampling locations cluster 3 mainly associated with B. sadleri, O. niloticus and Oreochr-
registered significantly higher Margalef indices than the inshore omis leucostictus (Figs. 3 and 4).
locations (Fig. 2). There was no significant difference in Shannon
diversity index between the inshore and offshore locations at the 3.4. Relationship between fish communities and environmental
reference site (H = 2.085 and H = 2.051) and Winday Bay variables
(H = 1.937 and H = 1.927). For Kirinya and Masese, the Shannon
diversity was significantly higher at the inshore sites with The canonical correspondence analysis (CCA) ordination
H = 2.085 and H = 2.012 than the offshore with H = 1.556 and showed a good relationship between fish species distribution and
H = 1.496 respectively. The offshore location of Murchison Bay reg- environmental variables as reflected by the high species-
istered a significantly higher Shannon diversity than the inshore. environment correlation coefficient associated with each axis
Hierarchical Cluster Analysis grouped the 10 sampling locations (Table 5). The most influential explanatory factors were electrical
into three clusters based on the similarity of fish species character- conductivity, total phosphorus, and total nitrogen concentration,
istics. The reference site locations (Lwanika inshore and offshore) secchi depth, and dissolved oxygen concentration. Most of the
and Winday Bay were grouped in the less polluted cluster 1 Murchison Bay inshore and offshore, and Kirinya inshore sampling
(Fig. 3). This cluster was mainly associated with haplochromine times (MBI, MBO and KI) were grouped closely and were associated
species which included P. redtail, Paralabidochromis crassilabris, with high nutrient levels and high electrical conductivity values. O.
H. victorianus and Paralabidochromis riponianus (Fig. 4). The med- leucostictus occurred under these conditions (Fig. 5). The reference
ium-polluted cluster 2 included the inshore and offshore sampling site (Lwanika, denoted by Li and LO) and Winday Bay (WI and WO)
locations of Masese and Kirinya (Fig. 3). This cluster was mainly were associated with high secchi depth readings, water tempera-
associated with T. zilli and A. alluaudi (Fig. 4). The Murchison Bay ture and dissolved oxygen concentrations. Most fish species
inshore and offshore locations were grouped in the highly polluted occurred under these conditions (Fig. 5). Total depth (shown by
924 I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928

Fig. 3. Hierarchical Cluster Analysis dendrogram showing three clusters of site composition, each cluster indicates sites with similar fish species composition. Homogeneity
within clusters was based on Euclidean distance and the heterogeneity between clusters was based on Wards method. Sampling locations L In and L Off = Lwanika Inshore
and Offshore, W In and W Off = Winday Bay Inshore and Offshore, MB In and MB Off = Murchison Bay Inshore and Offshore, M In and M Off = Masese Inshore and Offshore,
and K In and K Off = Kirinya Inshore and Offshore.

Fig. 4. PCA ordination biplot of the ten sampling locations based on the fish species variables for plot component one (X axis) and plot component two (Y axis). Note the three
groups (clusters) of the higly polluted Murchison bay, the medium polluted Kirinya and Masese and the least polluted reference site (Lwanika) and Winday bay.

Table 5
Summary of CCA axis length showing axis eigenvalues, correlation between species
4. Discussion
and the environmental gradients, and variance of species, following canonical
correspondence analysis of fish species abundance data in the study sites. 4.1. Composition and spatial diversity of fish species in the study sites
CCA axis 1 CCA axis 2 CCA axis 3
Eighty-three percent (83%) of the fish species observed during
Canonical eigenvalue 0.601 0.427 0.189
Cumulative percentage variance 20.2 34.6 41
the study were Cichlids. Cichlids are the most abundant and
Species environmental correlation 0.954 0.933 0.764 dominant group in Lake Victoria and other lacustrine environ-
Number of species (response 13 ments in the East African Great Lakes (Rosenberger and Chapman,
variable) 2000; Lundberg et al., 2000; Balirwa et al., 2003; Darwall et al.,
Number of environmental variables 12
2005; Witte et al., 2007). The other families were Alestiidae,
Total variance in species data 2.97
Centropomidae, Protopteridae and Mochokidae representing some
of the non-cichlid species inhabiting the lake (Namulemo, 2004).
the short arrow) did not influence the fish species distribution. The The inshore sampling location had the highest number of species.
first two CCA axes accounted for 34.6% of the total species variance This depicts the role of the wetlands, swamps, satellite lakes and
(Table 5) as explained in the CCA plot patterns (Fig. 5). shallow bays in the life cycle of many fishes, as well as the sites
I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928 925

Fig. 5. CCA plot of first and second CCA axes of fish species, environmental variables and their corresponding sampling stations. Eigenvalues: axis 1, 0.60, axis 2, 0.43. First 2
axes account for 34.6% of the variance. Monte Carlo test was significant (p < 0.001), 1000 permutations. Where sampling locations Li and LO = Lwanika Inshore and Offshore,
MBI and MBO = Murchison Bay inshore and offshore, WI and WO = Winday Bay inshore and offshore, MI and MO = Masese inshore and offshore, KI and KO = Kirinya inshore
and offshore. Numbers represent different sampling times.

serving as critical habitats for the recovering biodiversity in Lake the highly impacted effluent recipient Murchison Bay recorded
Victoria (Witte et al., 2000; Chapman et al., 2002, 2003; Balirwa the lowest indices. This suggests that urban wastewater discharges
et al., 2003). Overall, five major species were collected in all study that include industrial, domestic and storm-water effluent may
sites and these included O. niloticus, L. niloticus, P. redail, B. sadleri have negatively influenced the physicochemical variables conse-
and Tilapia zilli. Three of these species (O. niloticus, L. niloticus, and quently impacting negatively on the fish species diversity. Similar
T. zilli) were introduced, while B. sadleri and P. redtail are indige- trends of low species richness, diversity and dominance at the
nous, with O. niloticus being the most prevalent in all the studied highly impacted sites were observed from other poor quality lacus-
sites. These findings concur with earlier studies that reported the trine water bodies (Carol et al., 2006; Sekiranda, 2006; Ye et al.,
dominance of O. niloticus in Lake Victoria waters (Balirwa et al., 2006; Mondal et al., 2010). In the present study, water quality
2004a; Okaronon, 2004; Njiru et al., 2006, 2008). deterioration was highest in Murchison Bay followed by Kirinya
The dominance and wide distribution of O. niloticus compared and lowest at the reference site (Lwanika). Another observation
to other tilapiines, is mainly attributed to this species ability to was that, there was no significant difference in Shannon diversity
feed on a variety of food items, high fecundity, ability to grow to indices between the inshore and offshore sites of Lwanika and
a large size and capacity to survive in a wide range of physical Winday Bay. However, these indices were significantly higher at
and chemical conditions in the lake (Balirwa et al., 2004a; Okar- the inshore than the offshore in Kirinya and Masese while in Mur-
onon, 2004; Njiru et al., 2006, 2008). This could explain why O. chison Bay the offshore diversity was higher. The low Shannon
niloticus was not only the dominant species in this study but diversity indices at the inshore location for Murchison Bay could
was also observed in every site during each sampling time. be as a result of the poor water quality at the bays shoreline mak-
These findings agree with those of Sekiranda (2006) who ob- ing the environment less habitable by many fish species. This could
served O. niloticus, L. niloticus and B. sadleri among the major explain why a number of native Haplochromine species such as A.
species observed in three shallow bays of Lake Victoria. A similar nibula and A. latifasciata among others observed in the less im-
trend was observed in this study where by the other predomi- pacted sites were not observed in Murchison Bay. Brycinus jacksonii
nant species were L. niloticus, T. zilli and P. redtail. Lates niloti- was also not found in Murchison Bay. For species richness on the
cus was the most prevalent in Winday Bay offshore and the other hand, the Margalef richness indices were higher at the off-
second most prevalent in Lwanika, Masese and Kirinya offshore. shore locations except for the reference site. This implies that
Lates niloticus though a non-wetland species, the juveniles are although there was a higher diversity at the inshore, these sites
only found in shallow waters, usually close to the shoreline. had low species richness; while at the offshore locations, the diver-
Wetlands and shallow bays play a role in the life cycle of L. nil- sity was low but the species richness was high.
oticus probably by offering refuge to the juveniles from preda- Sampling locations were grouped into three clusters of least
tion. On the other hand the low frequency of L. niloticus in polluted, medium and highly polluted based on the fish species
Murchison Bay could be attributed to the low oxygen levels in characteristics. The highly polluted cluster was mainly associated
this bay. It is hypothesized that the high oxygen requirements with the highly adaptive and resilient O. niloticus together with B.
of L. niloticus limit its distribution. This explanation is supported sadleri and O. leucostictus; while the least polluted cluster was
by studies that recorded low abundance of L. niloticus in hypoxic mainly associated with haplochromine species which included P
environments (Chapman et al., 2002; Balirwa et al., 2003, 2004b; redtail, P. crassilabris, H. victorianus and P. riponianus. Haplochro-
Namulemo, 2004). mine cichlids and non-cichlid species in the Lake Victoria basin are
The reference site and Winday Bay that did not receive urban reported to have a relatively high tolerance to hypoxia
effluent discharges had the highest diversity and richness; while (Rosenberger and Chapman, 2000). However some species such
926 I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928

as H. victorianus was observed only in Lwanika and Windy Bay diversity and richness indicating that eutrophication reduced fish
while O. variabilis occurred only in Winday Bay and was among biodiversity. These findings agree with those of Carol et al.
the most abundant at this site. Oreochromis variabilis population (2006) who observed that trophic state indicated by chlorophyll
size is reported to be declining or to have disappeared in many or nutrient concentrations independently explained most of the
areas of Lake Victoria (Namulemo, 2004). This species was classi- variation of fish assemblages in 14 Spanish reservoirs. Here, abso-
fied as critically endangered by IUCN. Its occurrence in Winday lute and relative abundance of common carp was strongly related
Bay provides evidence about the resurgence of this species in some to the trophic state of the reservoir especially to total phosphorous
areas. Future studies in this bay could give a clear indication of the concentration.
status and prevalence of this fish in this site. This observation in Sites with poor water quality had low species richness and
Winday Bay could affirm the importance of some wetlands and abundance. Only one fish species, O. leucostictus seemed to be
shallow bays in the conservation of fish biodiversity and the resur- highly associated with and tolerant to eutrophication and as a re-
gence of some species that were considered extinct. sult thrived in the hyper-eutrophic Murchison Bay than other sites.
Another closely associated species that was also predominant in
4.2. Relationship between fish communities and environmental other sites was O. niloticus. These observations show that eutrophi-
variables cation and poor water quality had a negative impact on species
diversity. This is in agreement with reports that eutrophication
The distribution and abundance of fish species was influenced negatively impacts on fish biodiversity (Mugidde and Hecky,
by water transparency, dissolved oxygen concentration, electrical 2004; Witte et al., 2007; Hecky et al., 2009; Nyenje et al., 2010).
conductivity and nutrient levels in the water especially total These studies and the present study agree with the fact that there
phosphorus concentration, with more fish species associated is a relationship between urban effluent discharge and eutrophica-
with sites with high dissolved oxygen concentration, high secchi tion; and this compromizes the ecological integrity of surface
depth readings and low electrical conductivity levels. Similarly, waters, causing depletion of oxygen and negatively impacting on
distribution and abundance of fish species was influenced by fish populations and species diversity.
water quality, (Mrosso et al., 2004; Carol et al., 2006; Ye et al.,
2006; Mondal et al., 2010). The closest study conducted on the
5. Conclusions and recommendations
Tanzanian part of Lake Victoria revealed that fish species rich-
ness was significantly correlated with water transparency/secchi
In conclusion the distribution and abundances of the fish spe-
readings (Mrosso et al., 2004). The low species richness even in
cies was strongly influenced by water quality. The evidence pre-
rocky habitats was attributed to decreasing water transparency.
sented in this paper suggests that gradual anthropogenic
The present study and the findings by Mrosso et al. (2004), agree
changes due to the ongoing urbanization in the catchment such
with those of Seehausen et al. (1997) who indicated that water
as increased discharges of sewage, industrial and municipal
transparency/clarity is one of the strongest environmental pre-
wastewater into the environment has resulted into eutrophica-
dictors of fish species biodiversity especially in the haplochro-
tion which exerted its influence on water, nutrient levels, oxygen
mine species.
levels and electrical conductivity consequently affecting fish
It has been argued that the loss of water clarity causes loss of
diversity. Low diversities in the fish community structure were
genetic and ecological differentiation among haplochromine spe-
registered in the highly impacted sites as compared to the less
cies, and the decrease in water transparency associated with
impacted sites. Thus fish diversity may be a useful biological
eutrophication has probably contributed to the loss of species
indicator of water quality and this could be used in biomonitor-
diversity among cichlids (Seehausen et al., 1997; Mrosso et al.,
ing networks and programmes to assess water quality and in
2004). The authors associated the loss of transparency with
mapping out fish species hot-spot areas. Winday Bay should be
eutrophication. This association was also observed in this study,
treated as one of the sites relevant to conservation planning
as sites with high eutrophication were correlated with low sec-
for Lake Victorias indigenous fishes. This study also shows that
chi depth readings and low oxygen levels. Water transparency
the state of the aquatic environment as influenced by many soci-
is reported to be highly correlated with dissolved oxygen con-
etal activities in the catchments affects fish and therefore the
centration, as transparency enables light penetration which is re-
fisheries. The study further recommends that in order to pro-
quired in the photosynthesis process. In this study 11 out of the
mote the wise use and conservation of fish stocks, biodiversity
13 common occurrence species were strongly associated with
and water quality; conservation efforts should embrace changes
high secchi depth readings and high oxygen concentrations. This
in societal activities in the catchments that improve environ-
was predominantly obseved in the reference site and Winday
mental quality. For example promotion of treatment and biodeg-
Bay, the sites with significantly higher secchi readings and oxy-
radation of the effluents and sewage from point source polluters,
gen levels than Murchison Bay and Kirinya.
such as industries, factories and hotels before discharging into
Electrical conductivity was also reported to be an important
the lake.
variable that could predict the structure of fish assemblages in
addition to water transparency and other factors such as alkalinity
and macrophyte conditions (Ye et al., 2006; Mondal et al., 2010). Acknowledgements
Mondal et al. (2010) also reported that water depth influenced fish
diversity. However in the present study, water depth had no signif- The authors acknowledge the financial support received from
icant impact on fish species distribution. Similarly water depth had the Carnegie Research Grant through the School of Postgraduate
no significant effect on species diversity and distribution of the Studies, Makerere University, The Lake Victoria Research Initiative
small fishes in Niushan Lake (Ye et al., 2006). The present study (VicRes) and The Regional Initiative for Science and Education
and the findings in these studies (Mrosso et al., 2004; Carol et al., (RISE) through The Sub-Saharan Africa Water Resource Network
2006; Ye et al., 2006; Mondal et al., 2010), suggest that water (SSAWRN). This study would not have been possible without the
transparency and electrical conductivity are good environmental support of the director of the National Fisheries Resources Re-
predictors of species biodiversity in lacustrine habitats. The pres- search Institute (NaFIRRI), Jinja, Uganda, Dr. John Balirwa. Grati-
ent study also showed that sites with high nutrient levels espe- tude also goes to the coxswains, Mr. Charles Baliise and Mr.
cially total phosphorous concentration, registered low fish Mohammed Madoi, all at NaFIRRI.
I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928 927

References Kansiime, F., Keteyo, E., Oryem-Origa, H., Muchunguzi, P., 2007. Nutrient status and
retention in pristine and disturbed wetlands in Uganda: management
implications. Wetlands Ecol. Manage. 15 (6), 467543.
Arimoro, F., 2009. Impact of rubber effluent discharges on water quality and
Kayima, J., Kyakula, M., Komakech, W., Echimu, S.P., 2008. Study of the degree of
macroinvertebrate community assemblages in a forest stream in the Niger
Pollution in Nakivubo Channel, Kampala, Uganda. J. Appl. Sci. Environ. Manage.
delta. Chemosphere 77, 440449.
12 (2), 9398.
Bakema, R.J., Iyango, L., 2000. Engaging local users in the management of wetland
Kelderman, P., Kansiime, F., Tola, M.A., Van Dam, A., 2007. The role of sediments for
resources: The case of the National Wetlands Programme, Uganda. IUCN Easten
phosphorus retention in the Kirinyawetland (Uganda). Wetlands Ecol. Manage.
Africa Programme. Forest and Social Perspectives in Conservation. Working
15 (6), 481488.
Paper No. 3.
Lundberg, J.G., Kottelat, M., Smith, G.R., Stiassny, M.L.J., Gill, A.C., 2000. So many
Banadda, E.N., Kansiime, F., Kigobe, M., Kizza, M., Nhapi, I., 2009. Landuse-based
fishes, so little time: an overview of recent ichthyological discovery in
non-point source pollution: a threat to water quality in Murchison Bay, Uganda.
continental waters. Ann. Mo. Bot. Gard. 87, 2662.
Water Policy 11 (Suppl. 1), 94105.
Machiwa, P.K., 2003. Water quality management and sustainability: the experience
Balirwa, J.S., Chapman, C.A., Chapman, L.J., Cowx, I.G., Geheb, K., Kaufman, L., Lowe-
of Lake Victoria Environmental Management Project (LVEMP) Tanzania. Phys.
McConnell, R.H., Seehausen, O., Wanink, J.H., Welcomme, R.L., Witte, F., 2003.
Chem. Earth 28, 11111115.
Biodiversity and fishery sustainability in the Lake Victoria basin: an unexpected
Magurran, A.E., 2004. Measuring Biological Diversity. Blackwell Publishing, Oxford,
marriage? Bioscience 53 (8), 703715.
Balirwa, J.S., Nsega, M., Roijackers, R., Nkalubo, W., 2004a. Evolution of tilapia
Manly, B.F.J., 2005. Multivariate Statistical Methods: A Primer. Chapman and Hall/
fishery with specific reference to the Nile tilapia (Oreochromis niloticus Linn).
In: Balirwa, J.S., Muggide, R., Oguto-Ohwayo, R. (Eds.), Challenges for
Matagi, S.V., 2002. Some issues of environmental concern in Kampala, the capital
Management of Fishery Resources, Biodiversity and Environment of Lake
city of Uganda. Environ. Monit. Assess. 77, 121138.
Victoria. Technical Report No. 2. Fisheries Resources Research Institute, Jinja,
McGill, B.J., Etienne, R.S., Gray, J.S., Alonso, D., Anderson, M.J., Benecha, H.K.,
Uganda, pp. 7992.
Dornelas, M., Enquist, B.J., Green, J.L., He, F., Hurlbert, A.H., Magurran, A.E.,
Balirwa, J.S., Nsega, M., Sekiranda, S., 2004b. Wetlands and riparian zones as buffers
Marquet, P.A., Maurer, B.A., Ostling, A., Soykan, C.U., Ugland, K.I., White, E.P.,
and critical habitats for biotic communities in Lake Victoria. In: Balirwa, J.S.,
2007. Species abundance distributions: moving beyond single prediction
Muggide, R., Oguto-Ohwayo, R. (Eds.), Challenges for Management of Fishery
theories to integration within an ecological framework. Ecol. Lett. 10 (10),
Resources, Biodiversity and Environment of Lake Victoria. Technical Report No.
2. Fisheries Resources Research Institute, Jinja, Uganda, pp. 175186.
Mondal, D.K., Kaviraj, A., Saha, S., 2010. Water quality parameters and fish
Bierman, P., Lewis, M., Ostendorf, B., Tanner, J., 2011. A review of methods for
biodiversity indices as measures of ecological degradation: a case study in
analysing spatial and temporal patterns in coastal water quality. Ecol. Indic. 11
two floodplain lakes of India. J. Water Resour. Protect. 2, 8592.
(1), 103114.
Mora, C., Chittaro, P.M., Sale, P.F., Kritzer, J., Ludsin, S.A., 2003. Patterns and
Birungi, Z., Masola, B., Zaranyika, M.F., Naigaga, I., Marshall, B., 2007. Active
processes in reef fish diversity. Nature 421, 933936.
biomonitoring of trace heavy metals using fish (Oreochromis niloticus) as
Mora, C., Robertson, D.R., 2005. Causes of latitudinal gradients in species richness: a
bioindicator species: the case of Nakivubo wetland along Lake Victoria. Phys.
test with the endemic shorefishes of the tropical eastern pacific. Ecology 86,
Chem. Earth 32, 13501358.
Carol, J., Benejam, L., Alcaraj, C., Vila-Gispert, A., Zamora, L., Navarro, E.,
Mora, C., Tittensor, D.P., Myers, R.A., 2008. The completeness of taxonomic
Armengol, J., Garcia-Berthou, E., 2006. The effects of limnological features
inventories for describing the global diversity and distribution of marine
on fish assemblages of 14 Spanish reservoirs. Ecol. Freshwater Fish 15, 66
fishes. Proc. Roy. Soc. B 275, 149155.
Mrosso, H.D.J., Msuku, B.S., Seehausen, O., 2004. Relationship between water
Carr, G.M., Neary, J.P., 2006. Water Quality for Ecosystem and Human Health. A
transparency and species richness of surviving haplochromines in selected
Report by United Nations Environment Programme Global Environment
habitats in Mwanza gulf Lake Victoria. Tanzania J. Sci. 30 (1), 101108.
Monitoring System (GEMS)/Water Programme. Ontario, Canada, 132pp.
Mugidde, R., Hecky, R.E., 2004. Nutrient status and thermal stratification in Lake
Chapman, L.J., Chapman, C.A., Nordie, F.G., Rosenberger, A.E., 2002. Physiological
Victoria. In: Balirwa, J.S., Muggide, R., Oguto-Ohwayo, R. (Eds.), Challenges for
refugia: swamps, hypoxia tolerance and maintenance of fish diversity in the
Management of Fishery Resources, Biodiversity and Environment of Lake
Lake Victoria region. Comp. Biochem. Physiol. 133 (A), 421437.
Victoria. Technical Report No. 2. Fisheries Resources Research Institute, Jinja,
Chapman, L.J., Chapman, C.A., Schofield, P.J., Olowo, J.P., Kaufman, L.S., Seehausen,
Uganda, pp. 131144.
O., Ogutu-Ohwayo, R., 2003. Fish faunal resurgence in Lake Nabugabo, East
Mugisha, P., Kansiime, F., Mucunguzi, P., Kateyo, E., 2007. Wetland vegetation and
Africa. Conser. Biol. 17 (2), 500511.
nutrient retention in Nakivubo and Kirinya wetlands in the Lake Victoria basin
COWI and VKI, 1998. Kampala Water Quality Monitoring Programme: Murchison
of Uganda. Phys. Chem. Earth 32, 13591365.
Bay Water Quality Project, Report prepared for Ministry of Natural Resources
Munabi, C., Kansiime, F., Amel, A., 2009. Variation of water quality in Kakira
National Water and Sewerage Corporation by COWI in association with VKI,
catchment area, Jinja Uganda. Phys. Chem. Earth 34, 761766.
Kampala, Uganda.
Namulemo, G., 2004. Distribution of endangered native fish fauna and their
Crona, B.I., Rnnbck, P., Jidawi, N., Ochiewo, J., 2009. Murky water: analyzing risk
conservation in Lake Victoria. In: Balirwa, J.S., Muggide, R., Oguto-Ohwayo, R.
perception and stakeholder vulnerability related to sewage impacts in
(Eds.), Challenges for Management of Fishery Resources, Biodiversity and
mangroves of East Africa. Global Environ. Change 19, 227239.
Environment of Lake Victoria. Technical Report No. 2. Fisheries Resources
Das, S.K., Chakrabarty, D., 2007. The use of fish community structure as a measure of
Research Institute, Jinja, Uganda, pp. 109118.
ecological degradation: a case study in two tropical rivers of India. Biosystems
Njiru, M., Ojuok, J., Getabu, A., Jembe, T., Owili, M., Ngugi, C., 2008. Increasing
90 (1), 188196.
dominance of Nile tilapia, Oreochromis niloticus (L) in Lake Victoria, Kenya:
Darwall, W., Smith, K., Lowe, T., Vi, J.C., 2005. The status and distribution of
consequences for the Nile perch Lates niloticus (L) fishery. Aquat. Ecosyst. Health
freshwater biodiversity in Eastern Africa. Occasional Paper of the IUCN Species
Manage. 11 (1), 4249.
Survival Commission No. 31. IUCN, Gland, Switzerland and Cambridge, UK,
Njiru, M., Ojuok, J.E., Getabu, A., Muchiri, M., Cowx, I.G., Okeyo-Owuor, J.B., 2006.
Dominance of introduced Nile tilapia, Oreochromis niloticus (L.) in Lake Victoria:
Emerton, L., Iyango, L., Luwum, P., Malinga, A., 1998. The present economic value of
a case of changing biology and ecosystem. Proceedings of the 11th World Lakes
Nakivubo urban wetland, Uganda. IUCN Uganda National Wetland
Conference 2(2), 255262.
Programme Report. Wetlands Inspectorate Division, Ministry of Water Land
NWDR National Water Development Report, 2006. National Water Development
and Environment, 30pp.
Report: Uganda. 2nd UN World Water Development Report, Water, a Shared
FAO, 2006. National Aquaculture Sector Overview. Uganda. National Aquaculture
Responsibility, 197pp.
Sector Overview Fact Sheets. In: Mwanja, W.W., FAO Fisheries and Aquaculture
Nyenje, P.M., Foppen, J.W., Uhlenbrook, S., Kulabako, R., Muwanga, A., 2010.
Department. Rome. Updated 19 July 2005. [Cited 4 August 2010] <http://
Eutrophication and nutrient release in urban areas of Sub-Saharan Africaa>.
review. Sci. Total Environ. 408, 447455.
Gratwicke, B., 2004. Factors Affecting Fish Distribution in Coastal Habitats of the
Nzomo, R., 2005. Sustainable management of African Lakes the case of Lake
British Virgin Islands. Ph.D. Thesis, Oxford University, UK, 308pp.
Victoria. Paper Presented at the First Living Lakes African Regional Conference,
Gratwicke, B., Speight, M.R., 2005. The relationship between fish species richness,
2730 October, 2005. Imperial Hotel, Kisumu, Kenya.
abundance and habitat complexity in a range of shallow tropical marine
Odada, E.C., Olago, D.O., Kulindwa, K., Ntiba, M., Wandiga, S., 2004. Mitigation of
habitats. J. Fish Biol. 66, 650667.
environmental problems in Lake Victoria, East Africa: causal chain and policy
Hecky, R.E., Mugidde, R., Ramral, P.S., Talbot, M.R., Kling, W., 2009. Multiple
options analysis. Ambio 33 (1-2), 1323.
stressors cause rapid ecosystem change in Lake Victoria. Freshw. Biol. 55 (Suppl.
Oguttu, W.W., Bugenyi, F.W.B., Leuenberger, H., Wolf, M., Bachofen, R., 2008.
1), 124.
Pollution menacing Lake Victoria: quantification of point sources around Jinja
Henderson, P.A., Seaby, R.M.H., 1999. Community Analysis Package 1.52. Pisces
Town, Uganda. Water SA 34 (1), 8998.
Conservation Ltd., Lymington, UK.
Ogutu-Ohwayo, R., Balirwa, J.S., 2004. The history of fish communities, biodiversity
Henderson, P.A., Seaby, R.M.H., 2000. Environmental Community Analysis 1.33.
ad environment of Lake Victoria and lessons learnt. In: Balirwa, J.S., Muggide, R.,
Pisces Conservation Ltd., Lymington, UK.
Oguto-Ohwayo, R. (Eds.), Challenges for Management of Fishery Resources,
Henderson, P.A., Seaby, R.M.H., 2007. Community Analysis Package 4.0. Pisces
Biodiversity and Environment of Lake Victoria. Technical Report No. 2. Fisheries
Conservation Ltd., Lymington, UK.
Resources Research Institute, Jinja, Uganda, pp. 3137.
Jinja District Profile, 2003. Jinja District three-year development plan 2003/4, 2004/
Okaronon, J.O., 2004. Composition, biomass, distribution and population structure
5, 2005/6. Technical support of the district planning unit.
of the fish stocks. In: Balirwa, J.S., Muggide, R., Oguto-Ohwayo, R. (Eds.),
928 I. Naigaga et al. / Physics and Chemistry of the Earth 36 (2011) 918928

Challenges for Management of Fishery Resources, Biodiversity and Environment ter Braak, C.J.F., 1986. Canonical correspondence analysis: A new eigenvector
of Lake Victoria. Technical Report No. 2. Fisheries Resources research Institute, technique for multivariate direct gradient analysis. Ecology 67, 11671179.
Jinja, Uganda, pp. 4148. ter Braak, C.J.K., Smilauer, P., 2002. CANOCO reference manual and CANOCO Draw
Rosenberger, A.E., Chapman, L.J., 2000. Respiratory characters of three species of for windows users guide: software for canonical community ordination
haplochromine cichlids: implications for use of wetland refugia. J. Fish Biol. 57, (version 4.5). Microcomputer power, Ithaca, NY, USA, 500pp.
483501. Tokeshi, M., Arakaki, S., 2007. On estimating diversity in species-rich fish
Seaby, R.M., Henderson, P.A., 2000. Species Diversity and Richness Version 2.65. assemblages: an example with coral-reef fishes of eastern Indonesia. J. Fish
Pisces Conservation Ltd., Lymington, England. Biol. 71, 18791887.
Seaby, R.M., Henderson, P.A., 2006. Species Diversity and Richness Version 4. Pisces Witte, B.S., Msuku, B.S., Wanink, J.H., Seehause, O., Katunzi, E.F.B., Goundswaard,
Conservation Ltd., Lymington, England. P.C., Goldschmidt, T., 2000. Recovery of cichlid species in Lake Victoria: an
Seehausen, O., van Alphen, J.J.M., Witte, F., 1997. Cichlid fish diversity threatened by examination of factors leading to differential extinction. Rev. Fish Biol. Fisheries
eutrophication that curbs sexual selection. Science 277, 18081811. 10, 233241.
Sekiranda, S.B.K. 2006. Water Quality Patterns, Invertebrates and Fish Community Witte, F., Wanink, H.J., Kishe-Machumu, M., 2007. Species distinction and
Structure in Three Shallow Bays of Lake Victoria (Uganda) with Varying biodiversity crisis in Lake Victoria. Trans. Am. Fisheries Soc. 136, 11461159.
Catchment Land Use. Ph.D. Thesis, Makerere University, Kampala, Uganda, Ye, S., Li, Z., Lek-Ang, S., Feng, G., Lel, S., Cao, W., 2006. Community structure of small
411pp. fishes in a shallow macrophytic lake (Niushan Lake) along the middle reach of
Stainton, M.P., Capel, M.J., Armstrong, F.A., 1977. The chemical analysis of the Yangtze River, China. Aquat. Living Resour. 19, 349359.
freshwater. Canada Fisheries and Marine Service, second ed., Miscellaneous
Special Publications. Freshwater Institute, 180pp.
Swallow, B.M., Sang, J.K., Nyabenge, M., Bundotich, D.K., Duraiappah, A.K., Yatich,
T.B., 2009. Tradeoffs, synergies and traps among ecosystem services in the Lake
Victoria basin of East Africa. Environ. Sci. Policy 12, 505519.