Bulletin of the EANHS 29(3) 
16 
in all nine sites. In the six fragments differing in 
disturbance we measured understory vegetation 
structure as well as vertical vegetation density At the 
three control sites, we estimated tree density and 
disturbance 
Vegetation sampling was done by first selecting 
random points using a random numbers table These 
points were then plotted on a grid map of the site 
Fellows were divided into three teams of three people, 
and a fourth team of four people, including some local 
field assistants. Each team then selected a pacer 
recorder and measurer. The latter was also responsible 
for most of the equipment Each pacer from each team 
then calculated the length ol their pace Measurements 
to get to each random poim were then convened to 
number of paces For the first three teams, equipment 
included a compass to accurately reach the random 
point, lour long, tagged poles and .1 m rope to measure 
out 3x3 m plots Each team also carried 3 shun, ragged 
poles and I m rope to mark ofl 1x1 m plots, as well as 
a tape measure to measure the diameter of trees and 
Maexopsis at chest height. A tagged stake was also 
carried to mark the point at which the pacer veered 
from the forest path to get 10 a random point 
At each point, we marked off a lx I nt plot and a 
larger 3x3 m plot Percentage ground vegetation cover 
was then estimated at the smaller quadrant In the larger 
3x3 ni plots, the number of stems less than 30 cm in 
girth, as well stems greater than 30 cm in girth were 
counted To gel an indication ol disturbance, all 
Maesopsis less than 10 cm in girth were counted and 
diameter at chest height measured for those greater 
than 10 cm in girth in the larger quadrants 
The fourth team sampled vertical height density in 
all the study sites They measured the undcrsiory 
vegetation vertical cover less than 2 5mm height at 
each site ai random points and Irom a random angle 
using a vegetation profile board delineated in 5 m 
sections. 
Food Availability 
Forest interior species are thought to be particularly 
affected by forest fragmentation and disturbance One 
possible explanation for this is that forest disturbance 
affects food availability Wc sampled leal-litter 
invertebrate biomass from all nine foresl fragments. 
One fellow was assigned the task ol collecting leaf 
litter samples at thirty random points from each study 
site. These samples were then placed in Berliz funnels 
and then exposed to the sun This helped separate the 
invertebrates from the leaf litter Fellows then took 
turns in removing invertebrates specimens from the 
tunnels' collecting cup and sorting through the dried 
leaf litter for specimens thai had not (alien into the 
collecting cups. The specimens were then put in labelled 
plastic vials, with some preservative alcohol. These 
were to be oven-dried and weighed later, to estimate 
food availability. 
Microclimate 
A secondary effect of tropical forest fragmentation is 
that the microclimate and physical environment of the 
forest is altered !i has been hypothesised that many 
understory birds are adversely affected by forest 
fragmentation and disturbance because of changes in 
the forest microclimate Understory birds vary greatly 
in the distance that they are encountered from the t'oreM 
edge By monitoring temperature, relative humidity, 
light intensity and understory vegetation structure within 
the control sites and relating the distance with which a 
species is captured on average from the lores! edge, it 
should be possihlr to correlate these parameters with 
understory bird distribution 
At three sites that differed in terms of forest 
disturbance, fellows made five different simultaneous 
recordings ol temperature, relative humidity and 
luminescence Temperature and humidity were 
measured using a ihcrmohygromctcr and luminescence 
was measured using a lux meter These measurements 
were taken along four transects 1 5 m apart at 0 m. 20 m. 
40 m 100 m and 200 m from die forest edge at each 
site At each point along the transect, four 
measurements of each microclimatic parameter were 
recorded at chest height approximately I m apart in the 
four cardinal directions. 
KNOWLEDGE GAINED 
As a database co-ordinator and records officer, my work 
has been largely restricted to the office As a result my 
field experience was very limited. Earthwatch provided 
a perfect opportunity for me to participate in a field 
study, where I was able to learn many new field 
techniques The project also afforded me a unique 
opportunity to interact with fellow African professionals 
with a shared common interest in birds Though I had 
some prior bird ringing experience, it was mostly as a 
scribe At the I’samh are bird project. I learnt how to 
erect and put down mist nets, and observed how to 
remove birds from nets properly I also learnt the use 
or sequentially numbered poles, and uniform sized nets. 
Vegetation sampling. Icaf-litler invertebrate sorting 
and monitoring microclimate were all rew experiences 
The use of a random numbers table in the >cl«uon of 
random points for a study site was also a first I also 
Icarrn. sometimes painfully, how to reach a seemingly 
inaccessible random point! The project was an 
invaluable learning experience on bird monitoring field 
techniques. I hope to put these techniques to good use 
when I attend a similar study as part of a colleague's 
Ph D research on the globally threatened Taita Thrush 
Turdus ( olivaceous ) hellcrt This study is being 
conducted at a patchwork ol lorest fragments that nuke 
up the Taita 1 fills Forest Like the Usambaras. the Taita 
Hills are part of the larger range of Eastern Arc- 
Mountains. It Is an Important Bird Area (I BA) and one 
of Kenya's endangered tropical forest ecosystems 
I took photographs of some birds in the hand, which 
will be added to the departmental collection of bird 
photos Some of these birds were lifers, which have 
boosted my stagnating life list As I will encounter 
similar birds at Taita Hills, it is useful to have had 
