Looking fo r mangroves: detection o f the n o rth e rn m o st West-African
mangroves in a non-m angrove dom inated landscape using remote
sensing
Otero Viviana1, Katrien Q u isth o u d t2, Nico Koedam' and Farid Dahdouh-Guebas12
1 Laboratory o f Plant Biology and Nature Management, Mangrove Management Group, Faculty o f
Science, Vrije U niversiteit Brussel (VUB), Pleinlaan 2, B-1050 Brussels, Belgium
E-mail: voterofa@ vub.ac.be
2 Laboratory o f Systems Ecology and Resource Management, Faculté des Sciences, Université Libre
de Bruxelles (ULB), Avenue F.D. Roosevelt 50, B-1050 Brussels, Belgium
Detection o f m angrove forests in the Parc National du Banc d ’A rg uin (PNBA), M auritania, in two
satellite images is the overall objective o f th is study. These forests have an extrem ely discontinuous
d istrib u tio n and are o nly com posed o f Avicennia g erm inans (L.) Stearn. The mangroves do not
provide a direct social or econom ic benefit fo r the fisherm en villages located in the PNBA (DahdouhGuebas & Koedam, 2001). Nevertheless, m angroves provide breeding, spawning, hatching and
nursing areas fo r fish and shellfish, maintenance o f biodiversity, coastal protection, and aesthetic
value (Walters e t al., 2008). Furtherm ore, the mangroves o f the PNBA are interesting because o f
th e ir biogeographical characteristics. They are the n orthe rnm ost mangroves in West A frica and are
able to survive in a very dry environm ent (Dahdouh-Guebas and Koedam, 2001; Blasco, 2010). The
m angrove forests o f the PNBA have a poor regeneration rate and there are records o f a more
extended m angrove area 4,000 to 5,000 years ago (Mahé, 1985; Gowthorpe, 1993; DahdouhGuebas and Koedam, 2001). Remote sensing is a technique used to detect, describe, q u a n tify and
m o n ito r changes o f land cover and land-use patterns in spatial and tem poral dim ensions (DahdouhGuebas, 2002; Dahdouh-Guebas and Koedam, 2008). This study analyses tw o satellite images from
Cap T im iris area in the National Park. One image was captured in 2004 w ith the QuickBird Sensor
and the second image was captured in 2011 w ith the GeoEye-1 Sensor. The image o f 2004 has a
spatial resolution o f 61cm in the panchrom atic layer and 2.4m in the m ultispectral layers. The
image o f 2011 has a spatial resolution o f 41cm in the panchrom atic layer and 1.65m in the
m ultispectral layers. The detection o f mangroves in th is area is challenging because the reflectance
o f the surrounding land cover - silt, clay and desert soils, herbaceous te rre stria l vegetation, and
seagrass - confuses the detection process. In particular, th is ongoing study w ill: (i) v e rify w hether
given the current technical and local characteristics it is possible to detect mangroves
u nam biguously and (ii) id e n tify the more useful approaches to detect mangroves according to the
current characteristics. Both pixel-based and object-based classifications w ill be explored, the
fo rm e r being based on spectral inform a tion only, the latter on a com bination o f spectral and shape
features. The classifications w ill be evaluated by means o f an e rro r m atrix analysis.
References
Blasco F. 2010. A b o u t the mangroves o f Banc d ’A rguin, M auritania. ISME/GLOMIS Electronic Journal
8:13 - 15.
Dahdouh-Guebas F. 2002. The use o f remote sensing and GIS in the sustainable managem ent o f
trop ical coastal ecosystems. Environm ent, Developm ent and Sustainability 4:93 -112.
Dahdouh-Guebas F. and N. Koedam. 2001. Are the n orthe rnm ost mangroves o f W est-Africa viable?
A case study in Banc d ’A rguin National Park, M auritania. H ydrobiologia 458:241 - 253.
Dahdouh-Guebas F. and N. Koedam. 2008. Long-term retrospection on mangrove developm ent
using tran sdiscip lin ary approaches: A review. A quatic Botany 89:80 - 92.
Gowthorpe P. 1993. Une visite au Parc National du Banc d ’A rguin. Itinéraires - Présentation des
principales com posantes naturelles. Parc National du Banc d ’A rg uin, N ouakchott, M auritania:
193pp.
Mahé E. 1985. C o n trib u tio n à l’étude scientifique de la région du Banc d ’A rguin. Peuplements
avifaunistiques. PhD thesis. Université de Science et Technique du Languedoc, France: 576pp.
Walters B.B., P. Rönnbäck, J.M. Kovacs, B. Crona, S.A. Hussain, R. Badola, J.H. Primavera, E. Barbier,
and F. Dahdouh-Guebas. 2008. Ethnobiology, socio-econom ics and managem ent o f mangrove
forest: a review. A quatic Botany 89:220 - 236.
-
86
-