Symposium on Remote Sensing for Resources Development and Environmental Management / Enschede / August 1986
499
Photo-interpretation of wetland vegetation in the Lesser Antilles
B.Rollet
Office National des Forêts, Basse-Terre, Guadeloupe, France
ABSTRACT: The wetlands of Guadeloupe are typified through photo interpretation on panchromatic 1:20000 flown in
1984 and earlier. Examples of interpreted photos are given and discussed.
Geomorphology is a useful approach. Field observations for checking and collecting additional data especially
on soils are indispensable before producing an ecological map.
RESUME: Les terres basses de Guadeloupe sont classées en types de végétation par interprétation de photos
aériennes panchromatiques au 1:20000 prises en 1984 et antérieurement. On donne des exemples de photos inter
prétées et commentées. La géomorphologie est une approche utile mais des observations de terrain sont indis
pensables pour vérifier avant extrapolation et pour prendre d'autres données, en particulier sur les sols
afin d'établir une carte écologique.
1 INTRODUCTION
Wetland vegetation is understood in a broad sense
i.e. a vegetation under the influence of periodic
inundation (fresh or seawater) or with poor drainage.;
this includes mangroves, Pterocarpus and open swamps
with 0 to 10 ? salinity submitted to tides or perma
nent inundation during 6 to 8 months a year. Inun-
dable pasture lands are discarded.
2 WETLAND VEGETATION TYPES
The following is distinguishable on 1:20000 panchro
matic :
- Mangrove as a whole and its subtypes
- Pterocarpus swamp forest
- Dry deciduous or semi evergreen forest on karst
Chrysobalanus thicket, sedge and fern swamps
- Mudflat with algal cover, saltflats.
Minor types, essentially woodlands or thickets are
not readily recognizable : Hibiscus tiliaceus,
Thespesia populnea, Conocarpus erectus, Annona
glabra, Symphonia globulifera ; Dalbergia monetaria,
Montrichardia along rivers.
3 PHOTO INTERPRETATION
In Guadeloupe available airphotos are 1:20000 pan
chromatic flown by IGN in 1962-63, 1968-69, 1982,
1984, also 1:50000 in 1950; and on limited areas
1:10000 colour taken by ORSTOM. The same applies
to Martinique except that false colour 1:20000 was
flown is 1975 (chanteur et al, 1980).
The scale 1:50000 is indadequate to map distinct
forest types within mangroves but mature monospe
cific stands on large areas are recognizable (young
stages much less so). Rules of generalization are
necessary to map the innumerable mixtures of age-
classes and species in a readable way. The identi
fication of small areas e.g. 5 mm diameter of thickets
woodlands and forests will not be possible generally
for all the species enumerated above.
Detailed field checks are always necessary.
Cyperaceae (Cladium, Eleocharis, Rhynchospora) and
fern swamps (Acrostichum, Nephrolepis) are well
recognized as a whole but the species are difficult
to separate even with the help of ecological consi
derations. Unless there is a need for agriculture
(disease detection, yield prediction), it is unli
kely that special photography will be used for
forestry only.
Concerning the main vegetation types the situation
is much brighter. Pterocarpus forests are easy
to delimitate in general ; they show a dark velvety
tone when crowns are fused ; emergents, irregularly
open stands, transition zones, food crops beneath
the forest introduce differences in the normal
pattern. Large patches are quasi monospecific but
quite often there is an assemblage with scattered
rainforest species (Ficus, Symphonia, Cassipourea,
Maytenus guianensis), and sometimes with Rhizophora
(northern bank of the lower course of Grande Rivière
à Goyaves or Laguncularia (west of Rivière Salée),
in all cases undetectable.
Mangroves can be subdivided fairly easily when
stands are pure and mature : Rhizophora mangle
is almost black, Avicennia germinans is salt and
pepper (white with grey dots), Laguncularia racemosa
is darker than Avicennia. Young stages of the three
species are indistinct, pale grey, getting progres
sively darker for Rhizophora when stands grow older.
False colour is available in Martinique (not
in Guadeloupe). Chanteur et al (1980) mention that
Rhizophora and Laguncularia come out red, Avicennia
blue, Conocarpus blue grey. Young stages are not
distinguishable. These results confirm the findings
in Mexico (Rollet, 1974) except that on 1:10000
Conocarpus was rather purple grey.
On colour photos 1:10000 Avicennia is pale green,
more yellowish than the two other species, not
very distinct. On mature stands Rhizophora is bottle
green, Avicennia yellow green, Laguncularia grey
green easily mistaken for Rhizophora in mixtures.
Mudflats come out white or mauve. The use of colour
is questionable because of the extra cost for a
limited additional information.
False colour is an obvious breakthrough for spot
ting Avicennia pure or in mixtures. Another device
is needed to separate Laguncularia from Rhizophora.
Undergrowths are not seen in most cases so that
succession escape interpretation. Ill or decaying
trees turn blue in false colour, in particular
Rhizophora (Rollet, 1974), a source of misinterpre
tation. Other mistakes will be avoided with some
care : water bodies may be black or white in pan
chromatic according to sun reflection and should
not be taken for saltflats or mudflats. Film proces