ea Was at 
ithin low 
S at high 
changed 
5 ha of 
and built 
id cover 
ar) was 
> (129.2 
7), and 
esults 
Bemigisha, Jane 
For scenario2, (table 3) where open water was excluded, shrubland was found with the highest risk to papyrus (coefficient 
1), followed by agricultural fields (0.8), woodland (0.22), grassland (0.15) and built up area (0.11). The sensitivity analysis 
showed consistency as: Bu>Gra>wdl>Ag>Shr 
3.3.3 Risk potential to papyrus swamp combining the criteria in a GIS: 
Scenario 1 (Including open water) showed all the papyrus areas with potential risk levels ranging from medium to high. 
Most of the high risk areas were found to be in the north (see figure 2). 
  
Potential risk to papyrys swamp by other land cover types 
Scenario 1 (open water risk included) 
Legend 
0 Not considered 
1 Lew In scenario 2 , where open water was excluded, 
3 more papyrus area was found to be with low to 
ES a Moin medium classes of potential risk than in scenario 1. 
B About 543 ha of papyrus was found at low to 
: medium risk, which was 52 % of the total papyrus 
area, while the medium to high risk area covered 
502 ha, 48% of the papyrus area 
(table 4). 
  
— D 
  
0 High 
  
    
p 
hom 
  
Source: 
A weighted summation of the 
land cover types in order of importance: 
shrubiand, open water, Agricultural fields 
Woodland, Grassland anc BU up. 
by 
Jane Bemigisha, 1998 
10000 
  
  
  
Figure 2. Risk potential to Papyrus area by other landcover 
-M 
Area (ha) |%ge (ha) |%ge 
1 10 1044.9 100 
o2 ]542.9 2 
Table 4. Risk levels to Papyrus: Scenario 1 and 2 
  
4. DISCUSSION 
The results demonstrate a clear vegetation zonation from the lake shore to the adjoining higher elevation especially the one 
of 1967 that was similar to the one identified by Gaudet (1977). Interference from human activity was continuously 
increasing evidenced by agricultural fields and built up area by 1995. The zonation commonly started with the lake water, 
then papyrus swamp, but the subsequent classes were not in consistent order. In the different years, the order was exchanged 
amongst the different landcover types. Shrubland had the fastest change rate between 1984 to 1995, had the highest natural 
vegetation area, and followed papyrus in the zonation by the 1995. It also exhibited the highest total area change from 
Papyrus to natural vegetation, by the same year (table 2). 
The tendency for shrubland to dominate the process of overtaking papyrus, could be due to the indirect effect of climate and 
lake level or /and induced by human interference. Changes during climatic cycles have been found not to be accompanied by 
return to original state as a result of complete or almost complete dying off of species dominating at the beginning of the 
cycle or as a consequence of invasion of new more competitive species therefore leading to succession (Van de valk, and 
Welling, 1988). This fluctuation can be elaborated with the situation of flooding /inundation and drawdown which 
corresponded to the 1967/84 situation and 1984/1995 respectively. 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 171