502 
Figure 4. Canal Perrin 
Photo IGN 84 ANT 23200 341 Original 1:20000 reduced 
to about 1:27000 
1 Rhiz. bord de mer (fringe mangrove) ; 2 Rhiz. 
thicket with Avic. 10 m ; 3 Semi deciduous forest 
on limestone ; 4 Acrostichum ; 5 Tall Lagunc. around 
3 ; 6 Pterocarpus ; 7 Low prairie ; 8 Avic. 7 to 
10 m ; 9 Old canal colonized by mangrove ; 10 Cype- 
raceae ; 11 Fire sweep in Cyperaceae ; 12 Saltflat 
13 Cycloned Avic. + Lagunc. ; 14 Acrostichum thicket 
15 Young Rhiz. + scattered Avic. or Lagunc. ; 
16 Avic. thicket ; 17 Rhiz. thicket. 
Figure 5. Lamentin - Grande Rivière à Goyaves (G.R.G.) 
Photo IGN 63 ANT 036/200 159 Original 1:20000 reduced 
to about 1:27000 
1 Large Avic. ; 2 Avic. park-like : 3 Low prairie 
on high bank ; 4 Pterocairpus ; 5 Cyperaceae (was 
Pterocarpus before) ; 6 Chrysobalanus icaco thicket ; 
7 Nephrolepis swamp ; 8 Canefields ; 9 Acrostichum 
mixed with shrubs ; 10 Mixture Rhiz.-Avic.-Lagunc. 
The rare species Avicennia schauerana occurs at 
least in 5 sites in Guadeloupe : Le Moule, Fajou, 
N.E. Rivière Salée, Baie Mahault and Port Louis. 
It has been mentioned by Weiss on one site near 
la Caravelle (Martinique), also by Beard (Grenada) 
and Gooding (Barbados). Rhizophora harrisonii non 
uncommon on the Venezuelan coasts has not yet been 
spotted in the Lesser Antilles. 
8. COMMENTS ON MAPPING CONSTRAINTS 
The detailed legends of figures 2 to 7 give infor 
mation about local (punctual) interpretation, mainly 
on species identification, size of stands (height, 
crown texture) and mixtures of species. Narrow tran 
sects were cut in 1983 at random more or less perpen 
dicular to the eastern shore of G.C.S. eastwards. 
Transects 9 and 10 (Fig. 3) cut in dwarf Rhizophora 
or low stands are well detectable, the photo being 
taken soon afterwards (1984). 
Transects are extremely convenient owing to the 
possibility of a continuous description of vegetation 
along with its stereoscopie image, yielding photo 
patterns for comparison when the interpretation 
would be otherwise dubious. 
On the other hand geomorphology, inundation con 
ditions, soils can be examined and described more 
or less intensively according to the available time, 
allowing checks about the gradients of various fac 
tors : soil solution salinity at various depths, 
Q soil texture, proportion of peat and clay...). 
The wealth of details shown on panchromatic photos 
(various tones and structures of canopies) has to 
by reduced to a limited number of floristic-pedological 
units in harmony with the magnitude of areas to 
be mapped, scale of photos and final mapping scale. 
A 1:50000 (or smaller) scale may be quite sufficient 
to show landscapes and geomorphological units 
(karst outcrops, beaches, dunes, high and low banks, 
erosion or accretion zones...), large vegetation 
units (mangrove, swampforest, open inundable 
swamps...), broad soil groups (limestone, peat, 
clay, calcic or non calcic deposits...). 
A 1:10000 (or larger) scale is generally more 
suitable for forest management maps showing in detail 
the situation of the growing stock and growth condi 
tions i.e. forest sites and fertility classes. The 
trap to be avoided is to produce widely different 
patterns by the soil scientist on one side, the 
forester or botanist on the other side, barely raeca- 
ningful to each other. 
9. WETLAND MAPPING IN THE LESSER ANTILLES 
Stehld (1945-46, 1947) described the vegetation 
of the Lesser Antilles in many publications but 
produced only sketchmaps at small scales between 
1:300000 and 1:1200000. He regretfully introduced 
in the literature the ambiguous terms of mangrove 
maritime (for mangrove) and mangrove palustre (for 
Pterocarpus swamp forest).