Nadalahwa F. Madulu.

Reversed Migration Trends in the Kondoa Eroded Area 15

available resources in order to impro~" the quality of life of the people (URT
1992). Among the other policy goals, the issu~ of promoting sustainable
relationship between population, resources and the environment is given high
priority. The adoption of a NPP that recognises the linkages between population,
environment and development, is a positive step towards sustainable
development.
Various environmental and developmental projects, programs and policies aim at
environmental protection and improving the socio-economic and living
conditions of the people. Although such programs and projects are designed to
stimulate changes in the environment and socio-economic conditions of the
people, most of them have had impacts beyond the environment and development
spheres to affect the lifestyle of the people and their demographic behaviour as
well. Such tendencies have led to an increasing concern on the link between '
environmental factors and population. Given this complementary relationship
between the human and natural resources, any discussion of resource
mal1agement is incomplete if the demographic behaviours not closely examined.
This issue formed the bases for this study.

3. LINKAGES BETWEEN POPULATION, ENVIRONMENT AND

DEVELOPMENT
Various models have been developed to describe the linkages between .
popUlation, environment and development. These include the IPAT and PDE
models.
3.1 The IPAT Model
The IPAT model demonstrates the general relationShips between (I)mpact on
environment, (P)opulation, (A)ffluence and (T)echnology (UN 1993; UNFPA
1991; Ehrlich and Ehrlich, 1990). According to the Model, I = PAT where I is the
Impact on the environment; P represents the Population (absolute size, growth
rate, distribution); A reflects the Affluence (usually measured as GNP per
capita); and T demonstrates the impact of Technology (amount of pollution per
unit of GNP). Generally, the I=PAT formula mean that environmental Impact is a
joint function of Population x Affluence x Technology (UN 1993). According to
this model the people's Ufestyle, incomes and social organisations usually
determine consumption patterns. The technologies in use determine the extent to
which human activities damage or sustain the environment and the amount of
waste associated .with any level of consumption (UNFPA 1991). Population
determines how many persons there are and it acts as a multiplier that fixes the
total impact.). These three factors, compound each other's impacts. In other
words, for any given change in technology, any given level of consumption
patterns, any level of poverty, the more people there are, the greater is the overall
impact on the environment (UNFPA 1991). This equation yields insights of the
impacts that population growth and distribution can have on the quality and
quantity of critical natural resources.
16 Social Science Research Report Series, no. 20
.:,
3.1 The PDE Model
This model demonstrates that there are strong linkages between population,
development and the environment. Any change in one of these factors may
necessarily influence changes in the other two factors. The study on reversed
migration patterns in the Kondoa Eroded Areas (KEA) was b.ased on the logic of
the PDE Model that examine the linkages between population change (P), socio-
economic development (D) and environmental factors (E). Figure 1 demonstrates
these linkages.

Figure 1. The population, development and environment model

POPULATION ENVIRONMENT

DEVELOPMENT

The model aims at enhancing scientific understanding and demonstrating the \

long-term consequences of alternative policies on demographic, developmental II
and environmental spheres. It also gives a broad accounting framework that
spec;ifies the most important and immediate effects of one sector to another.
These linkages are demonstrated in figure 2, which shows that population and
development are entirely embedded in the environment.
In this study, (P)opulation was taken as the point of departure and considered to
be the basic driving force that, together with other factors, influence
(D)evelopment and the (E)nvironment. In this case, the reversed migration
component and other demographic factors reflect the P factor. The
(D)evelopment factor is reflected by the HADO project, which aims at
transforming people's lives, their socio-economic status, and the status of the
environment. The (E)nvironment factor reflects all environmental related impacts
in the KEA. The linlCages between these variables formed the ground for
analysing the impacts of reversed migration on the environment and development
in Kondoa District.
Nadalahwa F. Madulu. Reversed Migration Trends in the Kondoa Eroded Area 17

Figure 2. Basic Structure of the PDE Model

ENVIRONMENT

DEVELOPMENT
• Do~stic
Trade:
• International
EDeI'IY
by \'
Source ODd we Coasumpdon
by
Erooorr»c sector POPUiAnON
by
...
~,.

• AgeJGender/ Education!
Labour Force Ponicipatioo
• Chooses through
FettilitylMonalitylMigratio
nlSchoolioafl.abour Force
Erurv ODd Exists

GoverumnII policy
• Social
• EcoOOmlc
• EnviroDIIItDtal

SOURCE: Wolfgang (1995).

•
18 Social Science Research Report Series, no. 20

~
4. THE STUDY AREA .
. ,~

The study reported here was conducted in the HADO project areas in Kondoa
District (Figure 3). Kondoa District lies between latitudes 4°S and 50 S, and
between longitudes 34°E and 36~. The altitude range between 1000 and 2000
meters above sea level and the physical features are dominated by inselbergs and
undulating hills in the highlands and by flat and low lying plains in the eastern
and north-eastern parts. Broad sand-rivers caused by severe soil erosion in the
highlands are common in the lowlands and in the river valleys. The climate is
classified as semi-arid, with rainfall pattern ranging from 600 mm to 800 mm.
The vegetation cover consists of short grasses and scattered bushes. Human
activities such as expansion of farms, human settlements, forest clearing, bush
fires, and fuel wood have influenced the intensity and distribution patterns of the
vegetation (Mbegu and Mulenge 1984). Land use in the study area is typically of
subsistence nature. Small scale crop cultivation and livestock rearing are the
dominant economic activities. Land fallowing was used in the past as a method
of regenerating soil fertility. However, this method has been abandoned due to
land scarcity, severe soil erosion and population pressure.
The population of Kondoa District has more than doubled from 158,834 in 1957
to 340,554 in 1988. In 1978, about 49 percent of the population were children
under 15 years. In 1988, this age group had about 48 percent of population. An
age structure like this one is a typical example of high fertility and high growth
populations where the young age groups dominate.
During the colonial period, settlement patterns of Kondoa District were primarily
determined by measures taken during the Tsetse Flies eradication campaigns
(Madulu and Mung'ong'o 1990). Large areas of forests were cleared in order to
provide new lands for human habitation, cultivation and livestock gr~ing
(Mbegu and Mulenge 1984). The general pattern of the settlements during that
period was characterised by its scatteredness. Whereas the lowland areas were
sparsely populated, high concentration of settlement were in the Kondoa Irangi
Highlands especially around Lake Haubi. The major reasons for this
concentration were the availability of a reliable water source (i.e. Lake Haubi)
and favourable climatic conditions (Madulu and Mung'ong'o 1990). No major
changes in settlement patterns were recorded between 1960 and 1987.
In Kondoa District, the location and intensity of settlements is positively related
to the nature and severity of land degradation (Christiansson and Kikula 1996).
Similarly, the general distribution of settlement in KEA is highly associated with
the drainage patterns. In other words, the availability of water is among the basic
determinants of settlement distribution. For example, many homesteads are
concentrated arounQ Lake Haubi and along the major rivers such as Mkondoa,
Bubu and Mkuku (Figure 4).
 Nadalahwa F. Madulu. Reversed Migrati6n Trends in the Kondoa Eroded Area 19

4.1 Gubali Area ;.

./!!
This sub-village is also located in Kolo village. It is a very recent settlement that
developed in a protected area. Almost the whole of Gubali area was planted with
trees by the HADO project. This was one of the conservation measure to fight
severe soil erosion in the area. The first settler in Gubali was Mzee Mandela who
came from Masawi in 1985. Twelve other households moved into the area in
1992. Currently, the sub-village has abOut 62 households with a total population
of 296 persons. To demonstrate the rapid growth of the settlement, several social
.services have been established in the village centre. These services include a
primary school, shops and several tea rooms. The primary school was opened by
the District Commissioner in 1996 and had 100 pupils in 1998.
Although the district by-laws stipulate clearly that nobody is allowed to re~stablish
settlements in the areas from · where people were removed (HWK 1990), abe
resettlement process in Gubali is progressing without being checked. Nothing has
been done to stop the construction of new houses and opening of new farms in
Gubali as if the relevant environmental protection by-law is not applicable to the
area.
4.2 Choka Area
Choka is also located of Kolo ward. Its ·growth and expansion in a restricted area
has been due to return migration. During the 1970s, people from Choka were
moved to Kolo village. However, settlers in this area started to come during the
1988/89 season and since then, the speed of return has been high. The return
started by people establishing temporary shelters that were used for guiding crops
during the farming season. Slowly, these camps were turned into permanent
houses. At the moment the villages have a total number of 186 households,
distributed into four (4) sub-villages. These sub-villages are: Choka Centre (222
persons), Choka Asili (162 persons), Lusangi (144 persons) and Korokoso (142
persons). Already, an application has been submitted to the District Council to
upgrade the settlement into a full registered village that will combine Choka and
Gubali settlements together. If registered, the village will have 5 sub-villages,
namely: Choka Centre. Choka Asili, Lusangi, Korokoso and Gubali.
4.3 Kidongo Cheusi Area
Kidongo Cheusi is a sub-village of Loo village. Although settlements were ~!o
removed in this area, many households have come back. Similarly, thiys~b­
village has been an expansion area for the, main Loo village. The sub-village has
established as primary school of its own and several kiosks (shops) are emerging
at the village centre.
• 4~4 Mondo Area
The Mondo area covers the sub-villages of Old Mondo, Elele and Aya. All 'these
sub-villages belong to Mondo village. The extent of reversed migration in these
areas is worth noting because the resettlement process is occurring at the expense
20 Social Science Research Report Series, no. 20

of the past conservation measures. There is clear evidence ef r~nt resettlement

in Old Mondo and Aya. New houses are being built and many trees are being cut
for building and fuel purposes. Old buildings located at the Old Mondo
settlements are being rebuilt to enable new settlers to come (see Plates 4 and 5).
The sizes of settlements are rapidly expanding even to the edges of the gullies.
Almost everywhere, the expansion of settlements is accompanied by expansion
of farms.

Plate 5. Current settlement process: Rebuilding of houses at Old Mondo

.>
 "

Figure 3, Location of the study area in Kondoa District Z

UGANDA III
Q.
III
Dr
KENYA
',,~ -
..........
=-~
III
~

==
III
Q.

Shlnyongo. E.
F
~dOO
::0;,
Slngldo "'
.::

• • "'~
"'
~
Toboro
I~
•
\.

c)Q '
Dodomo
Morogoro .,.
":.. i3
TANZANIA • DoreBSoloom 1 ::l'
c:l'
;::
:=;l
Irlngo.
"'e-;::
Mbeyo.
s'
;;.
"'
~
KEY:
I •
Songeo
;::
~
l:l
t!l
...J
~ study Area ~
MOZAMBIQUE ~)..
~
l:l

• -
N
22 Social Science Research Report Series. no. 20
.:-

Figure 4. Settlement patterns in the Kondoa Irangi Highlands. Kondoa District (1987)

· - - · - - · Th. HAOO P,oj.ct bonda,y

_____ "'a'o. ,oads
- - - - .- -- Motor.b •• 'ractl'
.... _ _.... Vllla9. H.adqua,t.,s
;:.;.; .:.,..-:. Hou •• sl buildin9s .

, k ... .
Nadalahwa F. Madulu. Reversed Mig!ation Trends in the Kondoa Eroded Area 23

5. RESEARCH METHODS
The target population for this study was all households in the HADO project
areas. The emphasis was to explain how reversed migration patterns affect the
environmental conservation activities in the HADO project areas. The HADO
project areas include all villages in which the project has initiated some
environmental conservation activities. Most of these villages are located in the
Kondoa Eroded Area (KEA). Although the earlier plan was to extend interviews
to the peripheral villages surrounding the HADO areas, this was not possible due
to time and financial limitations. Thus, interviews were limited to 3 villages,
namely Loo, Kolo and Mondo. Within these villages, sample households were
selected from six sub-villages where the problem of reversed migration had
notable effects. Table 1 presents the sample villages, sub-villages and the •
distribution of respondents in each sub-village by sex.

Table I. Distribution of respondents by sex, village and sub-villages

Villages Sub-villages Males Females Total

Number Percent Number Percent Number Percent

Loo KlChekundu 28 78.4 8 2J.6 37 17.5

Mondo Aya 30 100.0 0 0.0 30 14.2
EleJe 20 55.6 16 44.4 36 17.0
Old Mondo 8 25 .0 24 75.0 32 15.1
Kolo Choka 24 6\.5 15 38.5 39 18.4
Gubali 30 78.9 8 2\.1 38 17.9
Total 141 66.5 71 33.5 212 100.0

SOURCE: Survey Data (1998).

The selection of the sample sub-villages was largely based on the nature of
environmental degradation in those areas, the type of conservation measures put
into I effect made in the past, and the prevalence of reversed migration. This
approach was important because most of the returning migrants and livestock are
actually concentrated in these sub-villages.
5.1 Sampling Procedures
Different sampling methods were used in selecting the households for the
interviews. The multi-dimensional approach used in the sampling process
involved the following steps. First, three villages were selected on the basis of
the intensity of environmental degradation and the HADO project activities. The
second step was the selection of sub-villages that were highly affected by
reversed migration. Third, lists of ten-cell leaders and heads of households were
compiled for each sub-village. To ensure coverage of all parts of the sub-village,
the selection procedures were based on each ten-cell leader from which sample
households were randomly selected. This purposeful select~ w/s intended to
24 Social Science Research Report Series. no. ·20
."
minimise bias and allow comparison of observations on the basis of population
pressure and environmental stress. In each sample village. at least one tenth of
the households were randomly selected for interviews.
5.2 Data Collection Procedures
Data was collected for a period of 3 weeks in August 1998. Emphasis was put on
establishing the extent of reversed migration trends and their impacts on
environmental conservation activities in the sample villages. Similarly, efforts
were made to document the important lessons that can be useful for the future
progress of the HADO project and in planning other related projects elsewhere.
Quantitative data was gathered through structured questionnaire, which was
administered to 212 villagers. This instrument was administered to family
members, preferably the heads of the household or wife. It was used to capture
general information of the households, socio-economic conditions, demographic
characteristics and migratory history of the household members. The age
distribution of the respondents in each sub-village is shown in table 2.
Table 2. Distribution of respondents by age and sub-village

RespondeOlls in each Sub-village Total

Age Groups
KlCheusi Ay. Elele Old Cook. GuboJi Percent Nwri>er
Mo ndo

20-24 S.4 20.0 2.6 S.3 3.3

2S-29 11.1 1.9 4

30-34 S.4 . 30.0 S.6 2S.0 12.8 28.9 17.S 37

3S-39 S.4 50.0 2.6 13.2 9.9 21

40-44 8. 1 20.0 19.4 .. IS.4 Il.2 12.3 26

4S-49 S.4 S.6 12.8 13.2 7.1 IS

SO-S4 18.9 19.4 2.6 IS.8 7.S 16

SS-S9 16.2 S.6 2S.0 8.0 17

60-64 16.2 11.1 10.3 2.6 7.1 IS

65+ 3S.1 16.7 50.0 41.0 7.9 2S.S 54

Total Percenl 100.0 100.0 100.0 100.0 100.0 100.0 100.0
1
Nurrber 37 30 36 32, 39 38 212

Mean Age S7.4 36.2 48.S SS.8 ,57.1 42.0 49.9

SOURCE: Survey Data (1998).

As it is observed from the table, the age of the respondents ranged between 20-24
and 65+ years. The majority of the respondent'S were, however, concentrated
between age 30-34 and 45-49. Variations existed between sub-villages, mainly
due to availability of the respondents at home during the interviews.
In addition to the actual structured interviews in the sub-villages, unstructured
interviews and discussions were made with the district and ward level leaders.
These interviews were vital in establishing and understanding the district position
Nadalahwa F. Madulu. Reversed Migration Trends in the Kondoa Eroded Area 25
.-
with regard to reversed migration' in the protected areas. A checklist was
prepared to ensure collection of uniform.information from all the key informants
in the study area. Discussions were held with key informants in order to capture
historical _migration trends and the impact of population pressure on the
environment in general and natural resource exploitation in particular. Lastly,
qualitative information such as the socio-economic activities and cultural
attitudes towards resource use were gathered through direct observations,
historical tales and from historical documents.
5.3 Data Analysis
The data were coded, entered in computer, processed and edited by using a SPSS
computer software. Frequency tables and cross-tabulation of variables were
performed. The observations from quantitative data were supplemented with
secondary information and qualitative data. The later was used to provide
background information by describing the social, cultural, economic and
environmental conditions and the implications of reversed migration in the sub-
villages and families.

6. PRESENTATION AND DISCUSSION OF RESEARCH

FINDINGS
6.1 The Resettlement Process in the Protected Areas
Between 1974 and 1976, thegovemment in collaboration with the district
administration and the HADO project, made a decision to resettle the population
from the severely degraded areas to the less degraded ones (Mbegu 1988, 1996).
This exercise was commonly referred to as "Operesheni Vijiji (Villagization)".
The resettlement exercise affected areas like Aya, Eelele, Old Mondo Choka,
Gubali; and Kidongo Cheusi. This exercise was followed by the destocking of
the whole KEA in 1979. The first phase of the HADO project put emphasis on
rehabilitation of the non-agricultural areas that were already severely degraded
with special emphasis put in the areas where people were removed after the
villagization exercise. The major activities that were done to reclaim the land
included construction of contour ridges, tree planting, planting of elephant
grasses in ' the sand rivers and sisal hedges on the contour ridges (Mbegu and
Mlenge 1984). In reCent years, a rapid resettlement process has been taking place
in many of the areas that were left uninhabited after the villagization and
destocking exercises. In most cases, the resettlement process is occuring at the
expense of all the environmental conservation efforts that were achieved -during
the past twenty five years. The demographic characteristics of the new
settlements are discussed in the following section.
6.2 Demographic Characteristics
The demographic characteristics of the new settlements indicate a rapidly
growing population. The total population of the 212 households visited was 981
people. The age distribution of this population is presented in table 3.
26 Social Science Rese~rch Report Series, no. 20
t

Table 3. Total population in the sample households by age and sub~~iIIages

Age People in all sample households Total

groups
KlCheusi Aya Elele Old Choka Gubali Percent NO.
Mondo

11.9 21.6 . 12.6 124

0-4 12.2 15.7 10.p 23.7 25 .0 15.3 150
5-9 12.2 13.0 12.9 3.7 8.5 11.4 112
10-14 17.4 13.5 14.4 3.7 4.5 8.4 82
15-19 7.6 18.9 13.4 7.7 8.9 4.5 9.5 95
20-24 6.4 11.9 11.4 23.1 4.4 2.3 6.9 68
25-29 4.7 3.8 8.5 23.1 5.9 11.4 8.1 79
30-34 4.1 11.9 5.0 3.0 5.7 4.8 47
35-39 4.7 8.1 5.0 6.7 4.0 4.1 40
40-44 2.9 3.2 5.5 3.7 5.1 2.5 25
45-49 5.2 1.0 3.7 2.8 3.4 33
50-54 4.7 3.5 7.7 1.7 2.2 22
55-59 5.2 1.0 7.7 3.0 0.6 1.7 17
60-64 3.5 3.0 17.8 2.3 8.9 87
65+ 9.3 5.5 30.8
Total Percent 100.0 100.0 100.0 100.0 100.0 100.0 100.0
Population 172 185 201 112 135 176 981
Households 37 30 36 32 39 38 212

SOURCE: Survey Data (1998).

As it has been demonstrated in the table, people in the young age groups
dominate the population in the study area. In almost all sub-villages, over 40% of
the populations comprised of children under 15 years of age. A High
concentration of the population in the young age groups is a characteristic of high
fertility populations.
Given the total population available from the sample households in each sub-
village, the average household size for the sample was 4.6 people per household.
However, the mean size of households was much higher in Aya (6.2) and Elele
(5.6) and the lowest in Old Mondo (3.5) and Choka (3.5). Table 4 shows the
distribution of households by sex.
Over half of the households have 4-6 persons. This may reflect both high fertility
and a strong extended family culture in the study area. These features reflect a
remarkable implication to resource exploitation in the new settlements because of
the built-in growth momentum of the population.
Nadalahwa F. Madulu. Reversed Migration Trends in the Kondoa Eroded Area 27
:, .,
Table 4. Household size by sub-villageS

Household Size Total

(Percentage of total households)
Sub-villages Mean
size
1-3 4-6 7-9 10+ Percent. No.

Kidongo Cheusi 32.4 48.6 13.5 · 5.4 100.0 37 4.6

Aya 56.7 43.3 100.0 30 6.2
Elele 19.4 38.9 41.7 H - 100.0 36 5.6
'it.
Old Mondo
Choka
25.0
51.3
75.0
43.0 5.1
--- 100.0
100.0
32
39
3.5
3.5
Gubali
Total (Percent)
15.0
25.0
16.3 5.3 2.6 100.0
100.0
38 4.6 •
56.1 17.5 1.4
Total 53 119 37 3 212 4.6
(Number)

SOURCE: Survey Data (1998).

6.3. Migration Patterns

Notable human and livestock population movements occurred during the early
1970s .. Most of the movements were initiated by the villagization exercise and, to
a certain extent, by the activities of the HADO project. The respondents were
asked to indicate the villages where they went afterthe resettlement in the 1970s.
Table 5 summarises the observations from the data on movements of population
in the KEA area.

Table 5. Distribution of respondents by previous dystination after the 1970's.

re-settlement.

Proportion of all people born Total

Sub-villages
Not Other sub- Other Other Other Percent Number
stated villages villages district region

Kidongo Cheusi 32.4 59.5 2.7 5.4 100.0 172

Aya 100.0 100.0 185
Elele 88.9 11.1 100.0 201
Old Mondo 100.0 100.0 112
Cholca 15.4 69.2 15.6 100.0 135
Gubali 15 .8 71.1 7.9 5.3 100.0 176
Total (Percent) 8.5 69.2 17.9 2.4 0.9 100.0
Total (Number) 18 149 38 5 2 212

SOURCE: Survey Data (1998).