April 1998 
3 
available on the number of crows in and around Mombasa. 
The aims of this study were to count the total number of 
crows that roost on Mombasa island and to assess which 
directions they arrived from. 
METHODS 
Our count was made on 28 August 1995, before the main 
breeding season which in Mombasa is well-defined, from 
September to January with a marked peak in October 
(Brown& Britain. 1980). From 16.30-I900p.ra., all crows 
arriving on the island from the surrounding mainland 
(north, west and south) were counted Counts were made 
by observers stationed strategically at four observation 
points around the island (see Fig 1): KMFR1 rooftop, 
Nyali Bridge, Makupa Causeway (Kibarani), and Likont 
Ferry (on top of cement silos). Each observation point 
was manned by two people, acting as observer and 
recorder respectively, the latter noting down the data on 
pre-designed forms. Birds that flew away from the island 
(t.e. in the opposite direction) were recorded separately. 
Since birds arrived in groups of varying size, the number 
of birds present in each group was recorded (although 
sometimes it was difficult to tell what constituted a single 
group). The crows were counted in 15 minute blocks. 
Numbers in each 15-minute block were recorded group 
by group, then summed. It was always relatively easy to 
make precise counts of total numbers, eliminating the 
need for rough estimates of group sizes. Counting was 
sometimes aided by the use of binoculars, but most counts 
were performed by naked eye The position of the 
different counting stations was chosen so as to minimise 
the risk of double counts. Some small overlap might 
have occurred of counts at KMFRI and Nyali Bridge, 
but this might be compensated by die gap (although 
narrow) between the reach of Nyali Bridge and that of 
Makupa Causeway. We therefore believe, that the counts 
are reasonably accurate, and the effect of possible double 
counts is considered small and insignificant. 
RESULTS 
In total, we counted 19,616 Indian House Crows crossing 
the water towards Mombasa island between 16.30 and 
19.00 h. Besides Indian House Crows, only two Pied 
Crows were observed during the counts. The majority 
of the crows (12,962 birds or 66. 1 %) arrived from the 
mainland north (8,512 birds at Nyali bridge plus 4.450 
birds at KMFRI), whereas 3.464 birds (17.6%) arrived 
from the west (Makupa) and 3,190 birds 
(16.3%) from the south (Likom). 
Figure 2 shows the distribution of the 
arrival of the birds over time. The peak 
of arrival was between 17.45 and 18.30 
p m Data for each individual counting 
station yielded similar temporal patterns. 
From this figure it becomes clear that 
the counting period covered nearly all the 
arrivals, and that the number of birds 
which arrived prior to 16.30p.m. or after 
19.00 p.m. (when it became too dark to 
observe anything) is likely lo be 
insignificant. 
Prior to crossing the water, the birds 
often appeared to gather in tree tops, roofs 
and electricity poles and wires at the 
waters edge. After their aggregations 
reached a certain size, birds started 
crossing the water as a large group, usually 
from 25 to 85 birds. This phenomenon was 
most noticeable during the peak period 
of arrival, between 17.45 and 18 30p.m. 
However, at least half of all our 
observations concerned groups of less 
than 20 individuals. Generally, the birds 
appeared to cross the water towards 
Mombasa island at sites where the water 
body was narrowest. 
A total of 150 crows were observed 
flying away from the island (i.e. 42 birds 
at Likom, 8 birds at KMFRI, 26 birds at 
Nyali bridge and 74 birds at Makupa/ 
Kibarani). In a few cases it could be 
established that these birds later flew back 
to Mombasa island joining other 
Figure 1. Map of the Mombasa area, showing the location of the 
counting stations mentioned in the text. 
