riterion 
ress the 
esulting 
sitivity 
; of the 
" found 
5). The 
eighted 
ange in 
ssland, 
Bemigisha, Jane 
  
3 RESULTS 
3.1 The distribution of Papyrus swamp and other land cover types in 1967, 1984, and 1995 
In 1967, the papyrus swamp was mainly concentrated in the northern shore of the main Lake Naivasha, as part of the flat 
lacustrine plain (Figure 1). There were fragmented clumps on the lake especially in the eastern part. The swamp covered 
about 1246 of the total land cover area . Other land cover types included open water covering the highest area, grassland, 
agricultural fields, shrubland (mostly Senna didymobotry, Polygonum spp , conyza spp, and sphaeranthus spp ) and 
woodland (mainly acacia spp) which dominated the natural vegetation. The 1984 papyrus swamp was still mainly 
concentrated in the north east flat lacustrine plain, but reduced 
  
   
  
  
Land cover 1967 to the river mouths of the Malewa and Gilgil. Fragmented 
dcover s clumps were found on the western side of the lake. The swamp 
Agric. field i area comprised only 5% of the total land cover area. In addition 
= Grassland 3E to the cover types observed in 1967, built up area was 
= as | identified, covering an area of 158.6 ha. In 1984 the lake level 
  
  
  
   
  
shrubland 
woodland 
was higher than in 1967. Of the natural vegetation cover, 
woodland still took the highest area , but was 1000 ha less than 
Topography boundaries. 
& 
    
in 1967. Agricultural fields increased in area about 2.5 times as 
large as in 1967. In 1995 the papyrus swamp was further 
reduced to an almost uniform fringe along the main Lake 
Naivasha, covering about 3% of the total land area (see also 
figure 2). Open water covered 13276.7 ha. Of the other natural 
vegetation, shrubland had the highest cover, followed by 
woodland and grassland. Agricultural fields increased by 
about 1500 ha, compared with 1984, while built up area 
increased by about 80 ha. 
Topography classes 
FL- Flat Lacustrine plain 
AFL- Almost Flat Lacustrine Plain 
U- Undulating ( plain to upper elevation) 
H- Hily 
  
  
Source: 
Visual interpretation of 
geval photographs 1967 
re s Scale 1:50,000 
WE HO me es 
ü 10000 by 
Jane Bemigishs, 1958 
  
  
  
Figure 1. Land cover maps for 1967 
3.2 Changes in the papyrus swamp cover from 1967-1984 and from 1984-1995: 
The papyrus swamp that remained without change for the period 1967-1984, comprised 851.25 ha, mainly concentrated 
along the North and North-west shore of the main lake. In 1984, 1959 ha of papyrus was lost to water area especially in the 
flat to almost flat areas. Of the natural vegetation, the largest part of the papyrus (304 ha) was converted into shrubland, 
followed by grassland (254.8 ha ), and woodland (182.3 ha), while 100.8 ha of papyrus was converted into agricultural fields. 
Some areas formally fringing the swamp were changed to papyrus especially the agricultural fields (104 ha), grasslands (94 
ha), and woodland (109 ha) on the north eastern shore. A small papyrus fringe of 156 ha remained unchanged between 1984 
and 1995. This was mainly located on the northern shore of the lake. Most of the fringe that was lost to open water in the 
period 1967-84 was regained (823 ha), while the floating clumps in the Western part of the lake changed again into water 
(164 ha). Most of the papyrus was converted into shrubland (555.25 ha), followed by agricultural fields (387.5 ha), woodland 
(195.3 ha) and only a minor area (8.8 ha) was converted into grassland. 
3.3 Potential risk to papyrus swamp area by other land cover types 
In this section, the three criteria and their quantification are presented, followed by the risk model results based on the two 
scenarios. The model inputs were derived from the status of the swamp between 1967 and 1995 (the results of the temporal 
analysis described in the previous sections). 
Criterion (1) : the distance of land cover types to the present papyrus: Papyrus risk index for each land cover type was 
shown by a value in the papyrus area representing the actual distance to the land cover type. The results are shown in table 1. 
The potential risk for papyrus swamp to be taken over by agricultural fields was found to be mainly in the northeast and 
southern areas with ranges of medium to high for the whole area. Built up area risk was found on the eastern papyrus 
swamp area with 23% of papyrus area at low to medium risk, while 77% was between medium and high risk. The open 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 169