Linking the Densities of Coral Associated Fish Functional
Groups to the Benthic Structure
Kennedy Osuka1’2*, Mare Kochzius1, Ann Vanreusel3, David Obura2and Melita Samoilys2
1Vrije Universiteit Brussels Plaine 2, 1050 Brussels, 2CORDIO East Africa #9 Kibaki Flats, Kenyatta Beach, Bamburi Beach P.O.BOX 10135 Mombasa 80101, Kenya
3 Marine Biology Section, Department of Biology, University of Gent, Ledeganckstraat 35, B-9000 Gent, Belgium
*E-mail: Kennedy.Osuka.Edeye@vub.ac.be or kennylardy5@yahoo.com
G roup
Introduction
a
R esem blance: S 17 Bray Curtis sim ilarity
c lu s te r groups
Understanding trophic interaction and the ecological status of
marine environment requires an assessment of benthic habitat,
fish abundance and biomass. Indeed there is a relationship
between benthic habitat and fish community (Pitman et al., 2007).
The taxonomic composition and ecological roles of fish functional
groups are diverse. A total of 12 coral associated fish functional
groups have been reported, namely piscivores, omnivores,
corallivores, invertivores, planktivores, detritivores, large
excavators, small excavators, scrapers, browsers, grazers and
grazers-detritivores. This study identifies benthic sub-habitats and
relate each sub-habitat to the densities of fish functional groups.
CO
_<D
Q_
E
03
CO
70
60
100
90
80
Sim ilarity
Figure 2: Cluster analysis of arcsine square-root transformed data of
benthic variables.
Hard coral "Soft coral "Fleshy algae "Turf algae "CCA
Rubble
60.0 i
50.0 -
Material and methods
Table 1 : Summary of objectives, method of data collection and data
analysis applied.
30.0 -
a>
Data analysis
Variables
Hard coral, soft Hierarchical
corals, fleshy
cluster analysis,
algae, turf
using SIMPROF
algae, crustose test, of arsine
coralline algae square-root
(CCA)and
transformed data
(Clarke et al. 2008).
rubble.
To link
Underwater
Fish species
Comparisons of
densities of
Visual Census of comprising 12
relative
abundance of
functional
fish functional selected fish
groups to the functional species groups
each functional
using 50 X 5 m
benthic sub­
groups across
habitats.
belt transect. 3 -7
benthic sub­
replicates per site.
habitats.
Method
Visual estimate of
percent cover of
benthic variables.
1-2 estimates per
site.
T
1
Q- 20.0
Objective
To determine
the benthic
sub-habitats
with similar
benthic
structure.
Ï
o 40.0 -
10.0
■
0.0
B
Cluster groups
Figure 3: Mean benthic cover by cluster groups (sub-habitats)
Variability of the relative abundance of functional groups
across the sub-habitats was observed, with only the piscivores
and invertivores, showing significant difference (p<0.05; Fig.
4). Piscivores were higher in abundance in turf algae
dominated sites (A) than in soft coral dominated sites (D)
while invertivores showed higher relative abundance in hard
coral dominated sites (C) than turf algae dominated sites (A).
Piscivores
Omnivores
Invertivores
Corallivores
o
CO
io
•1>
•1>
73
73
O
O
io
•i>
CD
o
£
O
73
>
<M
O
o
o
£
co
C
CM
■i>
Study area
>
■1>
cc
o
-
CC
B
This study uses benthic data
collected by D. Obura and reef fish
abundance data collected by M.
Samoilys, from a total 32 sites from
• Tanzania,
• Northern Mozambique,
• Comoros and
• Northeast Madagascar.
s
£
;
C
Tanzania fc*
>
4>
cc
O
%
—
<D
---1
---1
9y a
O
LO
_
io
-
o
-
ZD
0>
>
'II'
CC
o
CC
—
S ub -habitats (cluster g ro u p s )
S ub -habitats (cluster g ro u p s )
S ub -habitats (cluster g ro u p s )
Planktivores
Detritivores
Large excavators
Small excavators
&
o
CD
<
r>
o
o
73
■et
C
3
n
a
o
OJ
O
-lí>
73
C
S ub -habitat3 (cluster g ro u p s )
3
34>
o
CM
D
<
r>
o
73
•et
'Z.
ZD
-
O
&
•et
O
CO
Pn
O
OJ
a
<
r>
>
<
r>
cc
73
C
3
b
34>
>
CC
—I
LO
<D
■et
73
O
C
£
CO
OJ
%
<L>
O
^2 -
CO
o
-
□s5
A
>
22
o>
cc
io
-
o
-
a
n
BCD
B
V,.
■i
1
t■
•V
t i ^Comoros
S ub -habitats (cluster g ro u p s )
S ub -habitats (cluster g ro u p s )
Scrapers
Browsers
Grazers
G raze r-d etriti vo re s
&
*4
O
Ü
O
CM
B
73
C
fv
Jr
o
<
r>
>
io
.y
>1
/
o>
CC
o
<L>
O
-
□
S ub -habitats (cluster g ro u p s )
73
C
o>
>
4>
CC
&
io
-
<D
O
<D
Madagascar
Figure 1: Map of the study area
S ub -habitats (cluster g ro u p s )
Y
i)
$
r
Mozambique A»
S ub -habitats (cluster g ro u p s )
o
—
LO
OJ
io
—
a
B
73
C
C
LO
*
<D
>
_¡2
■li­
ce
OJ
<D
73
io
o
LQ
-
e □
o
_
LO
-
£ >
22
•li­
ce
□B
I-------1
S ub -habitats (cluster g ro u p s )
S ub -habitats (cluster g ro u p s )
S ub -habitats (cluster g ro u p s )
Figure 4: Relative abundance of 12 functional groups by sub-habitats
Conclusion
Results
Hierarchical cluster analysis of the benthic variables using SIMPROF
test revealed 5 distinct cluster groups A-E, henceforth called sub­
habitats (ANOSIM R = 0.769, p<0.001, Fig. 2). The sub-habitats
were characterized by dominance of different benthic variables: A
dominated by turf algae at 42.0 ± 6.0 % (mean ± se), B and C by
hard corals at 51.9 ± 3.3 % and 42.5 ± 3.2 % respectively, D by soft
corals at 33.3 ± 8.8 %, and E by fleshy algae - 37.5 ± 6.0 % (Fig. 3).
The study shows that on a broad scale, few functional groups
show preference to particular benthic sub-habitats. It is also
likely that biomass may show differences across the sub-habitats
and probably the broad scale visual estimates of the benthic
habitat structure do not fully explain the fish densities.
References
Clarke K.R., Somerfield RJ. and Gorley R.N. 2008. Testing of null hypotheses in exploratory community
analyses: similarity profiles and biota-environment linkage. Journal of Experimental Marine Biology and
Ecology, 366(1), 56-69.
Pittman S.J., Caldow C., Hile S.D. and Monaco M.E. 2007. Using seascape types to explain the spatial
patterns offish in the mangroves of SW Puerto Rico. Marine Ecology Progress Series, 348, 273-284.
Acknowledgement: Collection of data was supported by the Western Indian Ocean Marine Science Association (WIOMSA) through the Marine Science for Management
(MASMA) programme. This study was supported by VLIR-UOS.
O C E A N S
L A K E S
COR5 I9
East A fric a
^