Human Ecology

, Volume 34, Issue 6, pp 785–807

Urban Sheep Keeping in West Africa: Can Socioeconomic Household Profiles Explain Management and Productivity?

Authors

    • Department of Tropical Animal ProductionUniversity of Göttingen
    • Animal Production and Health DivisionFood and Agriculture Organization of the United Nations
  • Marianna Siegmund-Schultze
    • Department of Tropical Animal ProductionUniversity of Göttingen
  • Katrin Bednarz
    • Department of Tropical Animal ProductionUniversity of Göttingen
  • Simon Killanga
    • Institut de Recherche Zootechnique et Vétérinaire (IRZV)
Article

DOI: 10.1007/s10745-006-9011-7

Cite this article as:
Rischkowsky, B., Siegmund-Schultze, M., Bednarz, K. et al. Hum Ecol (2006) 34: 785. doi:10.1007/s10745-006-9011-7

Abstract

Rapid urbanization in Africa leads to a spatial concentration of people with different cultural origins and socioeconomic backgrounds resulting in a great diversity of life styles and livelihood strategies. One common strategy in Maroua/Cameroon and Bobo-Dioulasso/Burkina Faso is urban sheep keeping. Cluster analyses identified distinct socioeconomic groups with similarities between the towns: traditional livestock keepers, households headed by well educated government employees or traders, and more vulnerable groups formed of households headed by females, retired people or people with limited formal education. The household types in Bobo varied in their perception of the importance and the development of urban sheep keeping and their future plans. Those in Maroua differed in management intensity and in the potential to adapt their practices to the urban environment. Development interventions to reduce environmental pollution and risks for human health associated with urban sheep keeping need to account for these differences in the target group.

Keywords

urban sheep keepingtypologymanagement practicescluster analysis

INTRODUCTION

Between 1970 and 1995 African cities showed, at 5% p.a., the world's highest rate of urban growth (Delgado et al., 1999). Sub-Saharan countries reached average urbanization rates of 34% in 2000; some are heading for 50% (Tiffen, 2003). This rapid urbanization leads to a spatial concentration of people with different geographical and cultural origins, and with different religious, educational, and professional backgrounds. The combination of urban economic decline in the 1970s and 1980s and the continued growth of the urban population have led to an increasing reliance on livelihood strategies combining both cash income generation and subsistence production for direct consumption (Maxwell, 1996). The number of urban families engaged in agriculture increased dramatically during the 1980s (UNDP, 1996) and in West African cities sheep keeping became widespread.

Livestock keeping is not bound to the right to use a particular piece of land (Guendel, 2002), which makes urban cropping difficult and insecure (Dennery, 1996; Maxwell, 1996; UNDP, 1996). The other reasons for urban livestock farming include increasing household income, contributing to nutritional self-provision thereby saving cash expenses, and the fulfilment of sociocultural functions (Mougeot, 1994; Waters-Bayer, 1996). However, livestock farming in urban settings is often associated with a number of negative effects, e.g., dung on the sidewalk, clogged sewerage systems, traffic problems, contamination of water sources (Schiere et al., 2000; UNDP, 1996), and above all, human health risks (Mantovani, 2000). To minimize negative side-effects urban livestock farmers need to be adequately supported but in many African cities farming is technically illegal (Maxwell, 1995; Siegmund-Schultze et al., 1997) and urban livestock farmers have long been ignored by governmental and non-governmental organizations offering livestock services (Jahn et al., 1996).

A number of studies have shown that socioeconomic variables explain decisions for or against taking up urban agriculture in general and livestock keeping in particular. All income groups in developing countries are involved in urban agriculture (Smit, 1994), although families with a lower per capita income were more likely to keep animals in Khartoum (Richardson and Whitney, 1995). Egziabher (1994) reasoned that poor families were often large and tried to benefit from this labour resource. Indeed, household size significantly affected the probability of involvement in farming as a livelihood activity in Kampala, Uganda (Maxwell, 1995), and in Khartoum, Sudan (Richardson and Whitney, 1995). While vegetable gardening in Ouagadougou, Burkina Faso, was almost solely in the hands of those born in the city (Manshard, 1992), animal keeping in Khartoum was more likely in families with a rural background (Richardson and Whitney, 1995). Lee (1993) reported that about 60% of the families practicing urban agriculture in Addis Ababa had a higher level of education, while Freeman (1991) observed that the school education of the urban agriculturalists in Nairobi, Kenya, was only to primary level or absent entirely.

It seems plausible that if decisions for or against taking up crop or livestock production are related to socioeconomic characteristics, that this is also true for subsequent decisions on production objectives and management practices. Thus, a typology based on socioeconomic characteristics might lead to more meaningful results, which will enable the needs of urban livestock keepers to be addressed. Consequently, we hypothesised that (1) distinguishable socioeconomic groups existed among urban sheep keepers in West Africa; and (2) that these groups differed in their perceptions of sheep keeping, their management practices and in the productivity of their flocks. To test these hypotheses and to address the question whether socioeconomic criteria distinguishing sheep keepers in a certain town were unique or transferable to other towns, we compared the socioeconomic typologies of sheep keepers in two West African towns.

MATERIAL AND METHODS

Two case studies were conducted in the cities of Maroua, Cameroon, and Bobo-Dioulasso, Burkina Faso, from 1994 to 1998 as part of a research project funded by the Science and Technology for Development (STD3) programe of the European Community.

Data Collection in Bobo-Dioulasso

Bobo-Dioulasso is the second largest town in Burkina Faso. With over 300,000 inhabitants in 1996, it is located in the subhumid zone. The population density in the city districts ranged from 12 to 155 habitants ha−1. The majority of the economically active population worked in the tertiary sector (Ministère de l’Equipement, 1990). In 1995, a baseline study was conducted to characterize sheep production, which comprised 2096 urban and peri-urban households keeping small ruminants within a radius of 25 km from the center (Kocty-Thiombiano, 1999). Nine thousand and seven hundred sheep and 3300 goats were counted; the average flock size was four hair sheep and one goat, ranging from one to 47 small ruminants.

For the 1996 study two districts in Bobo were chosen that differed in terms of distance to the city center (central or peripheral), age of buildings and the year of parcelling the area (old or new), density of population and size of yards, infrastructure, and proximity of services. Thirty-seven sheep owners in the central district and 35 in the peripheral district were interviewed between April and June 1996. The sample was stratified to make it representative in terms of the main activities of household heads, according to the proportions recorded in the baseline sampling in 1995.

A structured questionnaire was administered to record socioeconomic household characteristics, which entered the cluster analysis (Table I). Furthermore, farmers were questioned regarding their circumstances and reasons for starting sheep husbandry, flock size, decision making and care taking, knowledge of livestock services, perception of the importance of sheep keeping, perception of the general prospects for urban sheep keeping, and their future plans.
Table I.

Description of Socioeconomic Variables used in the Cluster Analyses for Bobo-Dioulasso and Maroua

Socioeconomic variables

Categories in Bobo

Categories in Maroua

Nominal variables

 Keeping cattle

No, yes but not in town, 1–2 head in town, 3–20 head in town

Yes or no

 Cultivation of land

Yes or no

Yes or no

 City district

Central or peripheral

Central or peripheral

 Gender of HH head

 

Male or female

 Ethnic group

Mossi, Peulh, Mande group, foreigner

Pastoralists or other

 Religion

Muslim, Christian, Animist

Muslim, Christian, Animist

 Principal activity

Livestock keeping, crop cultivation, trading, governmental employee, housework, retired, othera

Livestock keeping, crop cultivation, trading, governmental employee, housework, othera

 Changed work

Yes or no

Not available

 No of HHs on yard

One, more than one

Not available

 Modern school education ofHH head

No school, Koran school, 1–6 years modern school, >6 years modern school

Yes or no

 Language knowledge of HHhead

Not applied, instead French literacy rate

Native language, native + other local language, French

 Modern school attendance offamily membersb

Nobody, 1–6 years (primary school), more than 6 years, university

Not available

 Principal activity of ancestors

Livestock keeping alone or + other, crop cultivation, crop cultivation + other except livestock, other

Not available

Metric variables

 Age of HH head

Quartiles: 17–39, 40–48, 49–60, 61—79 years

Quartiles: 22–40, 41–50, 51–57, 58–78 years

 Household size

Without small and school children, terciles: 1–3, 4–7, 8–34 persons

Quartiles: 1–5, 6–8, 9–12, 13–36 persons

 In town since

Quartiles: <17 years, 17–26, 27–36, 37–66 years, born in Bobo

Quartiles: <14 years, 14–20, 21–30.5, 31–60 years, born in Maroua

 French literacy ratec

Quartiles: 0–0.357, 0.358–0.625, 0.626–0.81, 0.811–1

Not applied, instead language knowledge of HH head

aOther principal activities include craftsman, musician, teacher, Muslim leader (“marabout”).bMaximum duration of school attendance of any member of the household.cHousehold members of all ages, who declared to be literate in French divided by members above 7-year-old assumed to be potentially literate.

Data Collection in Maroua

Maroua is the capital of the Far North province in the semiarid part of Cameroon. It had experienced a population growth rate of 5.6% from 1976 to 1987, and had more than 100,000 inhabitants in 1994. The city extends along a river and comprises about 30 districts, among which are former villages that became part of the city. The city districts were classified as central or peripheral according to their location and infrastructure. They differed in terms of access to pasture, distance to the main streets, and ethnic composition.

In November 1994 a rapid socioeconomic household survey was conducted comprising 102 sheep keepers, 54 in five central districts and 48 in 15 peripheral districts. Socioeconomic characteristics, cropping and livestock activities, as well as the main problems related to sheep keeping were recorded. These baseline data were used to select a representative subsample of 58 households in 18 districts for an in-depth study. From November 1994 to March 1995 additional data on socioeconomic characteristics and sheep management were collected. The current flock structure was recorded and the sheep owners were asked to recall all movements in and out of the flock during the preceding year (March 1994 to February 1995). From these data an annual flock production index was estimated. In addition, a seasonal flock production index was calculated on the basis of a herd survey for the period November to February, in a subsample of 38 flocks. The flock production indices (FPI) for the periods of one year and four months were calculated according to the method set out by Peacock (1987):
$${\rm FPI} = \frac{{\hbox{Sales (g) $+$ Slaughter (g) $+$ Social transaction out of flock (g)} \atop \hbox{$+$}\, \hbox{Net inventory change (g)}}}{\hbox{Initial flock size (kg)}}$$
where net inventory change = weight increase or loss + purchase – mortality – losses (thefts, etc.)

After socioeconomic groups had been determined, a randomly selected sample of households stratified by socioeconomic groups was revisited from June to September 1998, further to explore the differences in management characteristics. We had planned to visit a minimum of 15 households from each of the three groups, but only 36 could be retraced. Fifteen belonged to the first socioeconomic cluster (M-1), eight to the second cluster (M-2) and 13 to the third cluster (M-3). A detailed questionnaire covered flock size, husbandry practices, labor input and work-sharing among family members, motivation, development and future plans concerning sheep keeping in the household. Flock size was converted into ewe equivalents (32.5 kg liveweight) to allow a more meaningful comparison of labor time allocated to sheep production. Labor supply was calculated in adult male equivalents (AME). The differences between the clusters in frequencies for nominal variables were tested by chi2-analysis, the differences in metric variables by Tukey-test using the respective (Statistical Analysis System (SAS) for Windows, version 6.11) procedures.

Cluster Analysis

Cluster analysis was chosen as an analytic technique. When compared with logistic regression and correspondence analysis it allowed a more profound understanding of the socioeconomic determinants of the decision for or against sheep keeping (Siegmund-Schultze and Rischkowsky, 2001). Socioeconomic data from 56 households in Maroua and from 71 households in Bobo were used for the cluster analysis. The variables described the life circumstances and origin of the actors (age, religion, ethnic group, principal activity of household head, time spent in the city, household size, and city district), education (attendance of a modern school, level of education of household head), and agricultural activities (Table I). In Bobo additional variables relating to school education, the number of households per compound, the main activity of the family of origin, and whether or not the household head had changed his/her principal activity were included.

The cluster analysis demanded a consistent scaling of the entered variables. As the original scales of the variables differed, the lowest common level, i.e., the nominal scale, was used. Quartiles were calculated for metric variables and the households categorized accordingly. The case-driven approach applied in this study used hierarchical agglomerate clustering in order to classify livestock keepers into two or more nonoverlapping groups (proc cluster in SAS). The cluster procedure was performed in a number of steps. The first step was to compute a distance matrix using the Jaccard coefficient. As the variables did not show equal numbers of categories, it was necessary to perform separate steps for each variable to avoid overvaluing the more detailed ones. Then single distance matrices were added to complete the matrix, which was entered in the clustering process using Ward's minimum-variance method as the algorithm. The outcome was a set of cluster suggestions at different hierarchical levels. The evaluation consisted of a determination of the number of clusters (hierarchical level) and their characterization. The decision was based on the elbow diagram and the dendrogram as well as on R-squared statistics.

RESULTS

Bobo-Dioulasso

Socioeconomic Clusters of Sheep Keepers

The elbow diagram resulting from the cluster analysis of the sheep keepers in Bobo-Dioulasso suggested selecting the solution of 10 or 12 clusters, as the curve inflected at these solutions (Fig. 1). This, however, would have resulted in a very small number of cases per group, while the three cluster solution showed a disproportional distribution of cases (Fig. 2). As a compromise, the five group solution was chosen as it gave a sufficiently refined classification with each cluster containing at least ten households.
https://static-content.springer.com/image/art%3A10.1007%2Fs10745-006-9011-7/MediaObjects/10745_2006_9011_Fig1_HTML.gif
Fig. 1.

Elbow diagram of 71 sheep keeping households in Bobo-Dioulasso subjected to cluster analysis.

https://static-content.springer.com/image/art%3A10.1007%2Fs10745-006-9011-7/MediaObjects/10745_2006_9011_Fig2_HTML.gif
Fig. 2.

Dendrogram of 71 sheep keeping households in Bobo-Dioulasso (the circles show the chosen solution).

The clusters were then compared according to their socioeconomic characteristics:

Group B-1 (n=22). Two-thirds of the households were located in the central district. Three-quarters of the female-headed households were allocated to this group and constituted one-third of the group members. Forty percent of all retired household heads in the sample were placed in this group, accounting for one-quarter of the cases. The ethnic backgrounds were diverse, but all were Muslims. All were migrants having lived in Bobo for on average 32 years, which was a bit less than the average duration of stay of all migrants. One-third of the households shared their compound with others. The French literacy rate was lower than the average, and the maximum number of years of modern school education of a household member in this group was significantly lower than in the groups B-3 to B-5. The proportion of household heads who had attended Koran schools was at 50% quite high, while 30% had not attended any school.

Group B-2 (n=16). The households were located in similar proportions in the central and peripheral districts. The category of other (see Table I) principal activity occurred very frequently in this group. All were Muslims. The household heads, with an average age of 40 years, were relatively young and their age differed significantly from the other groups except B-4. One-quarter kept cattle. Almost all household heads born in Bobo were found in this cluster. The main activity of their ancestors was predominately neither cropping nor livestock keeping. The school education of the household heads varied, 44% had attended a modern school for between 1 and 6 years. However, the maximum number of years of modern school education of any household member was significantly lower than in the groups B-3 and B-4.

Group B-3 (n=10). All households were located in the central district. Fourty percent of the household heads were traders. Two-thirds of the Peulh in the total sample were present in this group, accounting for 60% of the group members. Eighty percent of the household heads belonged to the eldest age group and 20% to the youngest. Crop cultivation was not common, but 40% kept cattle in town. Half of the household heads came from families keeping livestock as their main activity. The number of economically active members in the households was higher than the average. The maximum years of modern school education of a member of the household was the highest among the five groups. The French literacy rate was also high and most of the household heads had a modern school education of a longer duration.

Group B-4 (n=11). All households lived in the peripheral district, one-third of household heads were retired and almost two-thirds were Mossi. Christians dominated this group, accounting for over 80% of group members. Half of the households cultivated crops; cattle keeping was rare. All were migrants, having lived a relatively short time in Bobo. All households had their own compound. The number of economically active members in the households was very small. The maximum duration of modern school education of household members was high and the French literacy rate was the highest, and significantly higher than in the other groups except B-3. However, the household head often had not attended school.

Group B-5 (n=12). Two-thirds were found in the peripheral district. Mossi dominated with three-quarters of the group members. Almost all were Muslims. All cultivated crops and more than half kept cattle, on average seven head. Group B-5 differed significantly from the other groups in terms of the area farmed and the number of cattle kept. There were a high number of economically active members in the households. Three-quarters of the household heads had already changed their job and the same proportion had not attended any school.

To summarize, the main characteristics of the groups were:
  • B-1: One-third female headed; many retired household heads; more than one household in the same compound; few household heads with a modern education; low French literacy rate.

  • B-2: Young household heads born in Bobo; principal activity of ancestors was not agriculture.

  • B-3: Young and old household heads; central district; Peulh; traders; principal activity of ancestors was livestock keeping; 40% kept cattle; higher educational level.

  • B-4: Recent Christian migrants in the peripheral district; one-third retired; half cultivated crops; small number of economically active persons; higher education levels of household members but not of household head.

  • B-5: Mossi; urban cattle keepers; having changed their jobs at least once; high number of economically active persons; three-quarters of household heads had no school education.

Differences Between Socioeconomic Clusters in Terms of the Characteristics of Sheep Keeping

The five clusters were compared for differences in the circumstances in which they started sheep keeping, their attitude towards sheep keeping, and their perception of it. The small number of cases often did not result in significant differences, but certain trends were apparent.

Age at starting sheep keeping and flock size. Household heads born in Bobo started sheep keeping significantly (p≤0.001) earlier (at 28 years) than the migrants (at 43 years), who started after they had already lived on average 19 years in town. Accordingly, the members of group B-2, who were mainly born in Bobo, started sheep keeping at an early age (Table II). The household heads of group B-3 arrived relatively late in town and started sheep keeping a long time after their arrival, when they were already quite old. However, their flocks were the second largest, though the duration of sheep keeping was the shortest of all groups. The migrant households in group B-5 started sheep keeping relatively early after arriving in Bobo. Thus, the duration of sheep keeping was quite long and their flocks were the largest.
Table II.

Age at Start and Duration of Sheep Keeping, Number of Sheep, Age at Arrival in Bobo and Years in Town before Starting Sheep Keeping in Bobo by Socioeconomic Group ‘‘B’’ (mean±SE)

 

Age at start of sheep keeping

Duration of sheep keeping

Number of sheep

Born in Bobo

Age at arrival in Bobo

 

Stay in town before starting

 

Group

Years

n

Years

n

Head

n

 n

Years

n

Years

n

B-1

42.1±2.6

20

11.8±2.7

20

6.5±1.1

22

0

21.0±1.9

19

20.6±3.7

17

B-2

29.9±2.6

16

10.1±2.5

16

6.9±0.8

16

13

19.0

3a

19.7

3a

B-3

49.8±4.8

9

7.9±1.8

10

9.5±1.5

10

1

27.9±4.2

8a

25.0±4.9

9a

B-4

40.3±3.4

10

10.0±2.8

11

6.5±1.5

11

0

24.1±2.0

10

16.4±3.0

11

B-5

37.0±3.8

8

15.8±2.6

10

10.7±1.7

12

0

24.5±2.6

10

12.8±3.6

10

aHousehold heads not born in Bobo.

Decision making and caring for sheep. In all groups the decisions concerning sheep keeping were almost always taken by the household head. In groups B-1 and B-4 the household heads also cared for the sheep. In remaining groups other family members were also involved.

Exchange of information. On average two-thirds of the sheep keepers were interested in questions concerning sheep management and they discussed them in the family but also with friends, and other sheep keepers at the livestock market. In group B-5 the proportion of households (83%) with this interest was highest, while in group B-2 this proportion was lowest (56%). Knowledge of persons or institutions (e.g., the Centre International de Recherche-Développement sur l’Élevage en zone Subhumide (CIRDES) at Bobo) offering information on livestock husbandry was highest in B-2 and B-3, where at least one could be named by each of the households. This percentage decreased to 73% in groups B-1 and B-4. Which institutions or persons were known also depended strongly on the sheep owner's location in the town.

Reasons for sheep keeping. When sheep keepers were asked for their reasons for starting to keep sheep in town, sociocultural aspects played the dominant role (45%), followed by economic reasons (23%), starting by chance (19%), and because of tradition (13%). Under sociocultural aspects the pleasure of having animals as company (‘C’est bon d’avoir des moutons à côté’) and the benefit of keeping certain household members occupied were subsumed. The ‘by chance’ category refers to instances where sheep keeping was started due to unforeseen circumstances such as inheritance or getting sheep as a present (more than 50% of the women started for this reason). Economics were important as a reason for starting sheep keeping for 33% in group B-1 and 50% of the households in B-5.

Future plans concerning sheep keeping. In group B-1 more than 50% envisaged improvements in sheep management (Table III). This was the case for one-third of the households in B-5, while in the other groups only around one-quarter named this option. In group B-1 and group B-5 none of the households would be interested in investing in anything other than sheep.
Table III.

Future Plans of Households in Bobo concerning Sheep Keeping by Socioeconomic Group

 

% Households

Future plans

B-1

B-2

B-3

B-4

B-5

No plans, just maintain

21

31

33

25

25

Increase flock size

16

25

22

13

25

Improve feeding, breeding, build a shed

53

25

22

25

33

Continue outside the town

11

13

11

13

17

Invest in something else/other livestock

 0

6

11

25

 0

Note. The sum of the columns does not always add up to 100% due to rounding up. The numbers in bold are those percentages which are of special interest to the analysis.

Opinion concerning the future of sheep keeping in town. The majority of households in groups B-1, B-2, B-5, and to a lesser extent B-4 expected a positive development, while group B-3 predicted a decreasing importance. Also, households in group B-3 declared that their children would not continue sheep keeping (40%) or that they would decide about it later (60%). In the other groups at least 40 % (the smallest percentage being in B-4) were sure that their children would continue the activity.

Maroua

Socioeconomic Clusters of Sheep Keepers

The elbow diagram (Fig. 3) resulting from the cluster analysis of 56 sheep keepers in Maroua suggested choosing the cluster solution of six groups, as the curve inflects at this solution. But the three group solution (Fig. 4) already offered a considerable reduction in distance between the clusters, which was not much improved by the six cluster solution. The grouping with three clusters also gave a sufficiently large number of cases per group and a meaningful typology.
https://static-content.springer.com/image/art%3A10.1007%2Fs10745-006-9011-7/MediaObjects/10745_2006_9011_Fig3_HTML.gif
Fig. 3.

Elbow diagram of 56 sheep keeping households in Maroua.

https://static-content.springer.com/image/art%3A10.1007%2Fs10745-006-9011-7/MediaObjects/10745_2006_9011_Fig4_HTML.gif
Fig. 4.

Dendrogram of 56 sheep keeping households in Maroua (the circles show the chosen solution).

The three groups differed significantly (p<0.05) in terms of age, ethnic group, religion, and principal activity of the household head, whether or not the household head was born in Maroua and the duration of his/her stay in town, location in a central or peripheral city district, the percentage of female headed households, the percentage of household heads with a modern school education, and the percentage practicing cattle keeping; a weakly significant difference (p<0.10) was found in terms of household size (Table IV). The percentage practicing cropping in town did not differ. However, the average size of cultivated land was larger in group M-3 (12.3±7.3 ha) than in group M-2 (3.4±2.0 ha), and smallest in group M-1 (1.2±0.21 ha).
Table IV.

Frequency (%) of Households with Certain Socioeconomic Characteristics by Socioeconomic Group ‘‘M’’ in Maroua

Socioeconomic characteristics

Group M-1

Group M-2

Group M-3

χ2 Significance

No. of samples

25

12

19

 

Age of household head

   

0.03a

 22–40 Years

24.0

75.0

21.1

 

 41–50 Years

32.0

25.0

26.4

 

 51–57 Years

16.0

0

15.8

 

 58–78 Years

28.0

0

36.8

 

Household (HH) size

   

0.08a

 1–5 Persons

36.0

8.3

26.3

 

 6–8 Persons

40.0

16.7

15.8

 

 9–12 Persons

16.0

33.3

26.3

 

 13–36 Persons

8.0

41.7

31.6

 

Years in town

   

0.001a

 Born in Maroua

8.0

16.7

79.0

 

 <14 Years

8.0

33.3

5.3

 

 14–20 Years

32.0

25.0

5.3

 

 21–30.5 Years

16.0

25.0

5.3

 

 31–60 Years

36.0

0

5.3

 

City district

   

0.03

 Central

52.0

25.0

15.8

 

 Peripheral

48.0

75.0

84.2

 

Principal activity

   

0.001a

 Livestock keeping

28.0

16.7

21.1

 

 Cropping

24.0

16.7

47.4

 

 Trading

12.0

0

15.8

 

 Governmentalemployee

0

66.7

0

 

 Housework

4.0

0

10.5

 

 Other

32.0

0

5.2

 

Belonging to pastoralist ethnic group

8.0

8.3

89.5

0.001

Female headed households

40.0

8.3

10.5

0.03a

Religion

   

0.001a

 Animist

36.0

16.7

0

 

 Christian

32.0

66.7

0

 

 Muslim

32.0

16.7

100

 

HH heads with modern school education

0

100

94.8

0.001a

Keeping cattle

4.0

16.7

68.4

0.001

aThe results of the Chi2 test have to be treated with care due to small group numbers.

Based on these differences the cluster groups can be characterized as follows:

Group M-1 (n=25). This group consisted of migrants, more than half of whom had already lived for more than 27 years in Maroua; 92% of the households were smaller than 13 persons and, thus, on average were smaller than M-2 and M-3. The proportion of female headed households was very high; ten out of the total 13 female headed households were in this group. Except for one household, no cattle were kept and none of the household heads had attended a modern school. At a higher level of homogeneity the group was composed of two subgroups of almost similar size. One subgroup (n=13) contained the ten female headed households with small household sizes. Two-thirds of these household heads were animists; most of the animists (73%) in the sample were allocated to this subgroup. The household heads of the second subgroup (n=12) carried out different principal activities and the households were of an average size (10.3 persons).

Group M-2 (n=12). Three-quarters of the household heads were between 22 and 40-years-old (the youngest quartile); none of the household heads was older than 50 years. Most of the households were migrants, having passed less time in Maroua than the first group. Two-thirds were Christians. Most of the households were of a large size (75% were larger than 13 persons) and three-quarters lived in peripheral districts. All governmental employees were in this group, which represented two-thirds of all group members. Only two households kept cattle. All household heads had attended a modern school. The random sample of revisited households in this cluster showed a significantly higher proportion of educated household members (80%) than the other two groups (38%). Forty-three percent of household heads had at least once changed their main activity in comparison to 13 and 23% in the other groups.

Group M-3 (n=19). More than three-quarters were born in town; most of them lived in peripheral districts. Livestock keeping or crop cultivation was the main activity of more than two-thirds (68%) of the group and half of the traders were found in this group. Only one group member did not cultivate crops as his either first or second most important activity. All were Muslims. Ninety percent of the household heads belonged to an ethnic group who are traditionally livestock keepers and 68% kept cattle. Except for one, all household heads had not attended a modern school. The main activity (61% of all cases) of their ancestors was, indeed, livestock keeping, either alone or in combination with cropping, while this was the case for no more than 25 % of households in the other two groups.

To summarize, the three groups were comprised of:
  • M-1: Migrants not keeping any cattle and generally without school education; part of this group were small, female headed households and the other part were diverse households with different main activities.

  • M-2 Younger household heads often with larger households and generally a good school education.

  • M-3 Traditional cattle and sheep keepers, mostly already older and without a school education.

Differences Between Socioeconomic Clusters in Characteristics of Sheep Keeping

The three socioeconomic clusters were compared in terms of sheep keeping characteristics, such as husbandry practices, their future plans, flock productivity, and labor invested in sheep keeping (Table V).
Table V.

Frequency of Households (%) Duration of Sheep Keeping, Sheep Husbandry Practices, Knowledge of Organizations, Combination of Livestock, Future Plans Concerning Sheep Keeping and Participation in a Savings Club by Socioeconomic Group ‘‘M’’ in Maroua

Characteristics of sheep keeping

M-1

M-2

M-3

χ2 signficancea

No. of samples

25

12

19

 

Keeping sheep since

   

*

 <7 Years

44.0

50.0

11.8

 

 7–13 Years

16.0

25.0

11.8

 

 13–22.5 Years

24.0

25.0

35.3

 

 >22.5 Years

16.0

0

41.2

 

Overnight location

   

*

 Yard

12.0

16.7

31.6

 

 House

44.0

41.7

5.3

 

 Shed

44.0

41.7

63.2

 

Buying commercial feed

32.0

83.3

84.2

Trimming hooves

64.0

50.0

26.3

**

Knowing governmental organisations

68.0

100.0

73.7

*

No. of samples

15

8

13

 

Main treatment of diseases

    

 None

14.3

0.0

0.0

 

 Traditional

28.6

14.3

30.7

 

 Drugs

42.9

42.9

69.2

 

 Assistance of a veterinarian

14.3

14.3

0

 

 Slaughter

0

28.6

0

*

Knowledge of traditional medicine

   

**

 None

53.3

71.4

7.7

 

 Knowledge

6.7

0

30.7

 

 Knowledge of application

40.0

28.6

61.5

 

Experienced sheep mortality

93.3

50.0

84.6

**

Feeding system

   

n.s

 Grazing throughout

15.4

57.1

41.7

 

 Scavenging

30.7

14.3

8.3

 

 Stall feeding

7.7

14.3

8.3

 

 Pasture r.s./scavenging d.s.b

23.1

0.0

25.0

 

 Pasture r.s./yard d.s

23.1

14.3

16.7

 

Part of saving club

   

**

 No

53.3

37.5

81.8

 

 One family member

26.7

0

0

 

 More family members

20.0

62.5

9.1

 

Future plans

   

*

 None

6.7

25.0

0

 

 Increase sheep flocks

93.3

50.0

92.3

 

 Other plans

0.0

25.0

7.7

 

aLevels of significance: n.s.: not significant.br.s.: rainy season, d.s.: dry season.p < 0.10; *p < 0.05; **p < 0.01; ***p < 0.001.

Duration of sheep keeping. The households in M-3 had the longest experience of sheep keeping: 77% of the households had kept sheep for longer than 13 years, while this was the case for 40% in M-1 and 25% in M-2.

Husbandry practices. More than half of the households in M-1 allowed scavenging of their sheep for at least part of the year, which was a (non-significantly) higher proportion than in the other two groups. The percentage of households grazing their animals throughout the year was highest in M-2. Only 32% of the households in M-1 bought commercial feed in comparison to more than 83% in M-2 and M-3. In the subcluster dominated by female headed households (n=13) within M-1 this percentage was, at 23%, even lower. Only one-quarter of the households in M-3 trimmed the hooves of their sheep compared to more than half in the other groups. Knowledge about traditional medicine and how to apply it was very frequent in M-3 and least common in M-2. The main treatment in case of diseases in all groups was the application of veterinary drugs, with the highest frequency in M-3. Fourteen percent in M-1 and M-2 called for a veterinarian. In M-2, 29% slaughtered their sheep in case of disease, which was not done in the other groups. Less than half of the households in M-1 and M-2 owned a shed, while this was the case for two-thirds in M-3. M-3 households only rarely admitted their sheep overnight into the house, while more than 40% in M-1 and M-2 shared their house overnight with their sheep. While 85% or more of the households in M-1 and M-3 reported deaths of sheep in the past year, this was the case for only 50% in M-2.

Knowledge of organizations offering services which support livestock husbandry. The households in M-2 had the best knowledge of such organizations: all household heads knew at least one governmental organization in comparison with 74% in the other groups; 92% knew a semigovernmental organization compared to less than 79% of the others; and 33% knew of NGOs, compared with less than 24% of the others. The latter two differences were not significant. All households in M-3 and 93% in M-1 would join a cooperative to receive credits, subsidies, and to be able to buy cheaper feeds or in expectation of a combination of these benefits; in M-2 this interest was smaller (75%). Twenty-nine percent in M-1 also expected an exchange of information from joining cooperatives; this was not an expectation in the other groups.

Future plans and reasons for sheep keeping. Ninety-three percent in M-1 would like to increase their flock size compared to only 50% in M-2. Only two-thirds of the households in M-1 believed that their children would continue sheep keeping, while this percentage was, at about 85%, higher in the other two groups. The groups did not differ in their perception of the future of sheep keeping in Maroua. Seventy-five percent of the households believed that the number of households involved in urban sheep keeping would increase or be maintained, 11% thought it would decrease, and none believed that the activity would disappear. Also there were only small differences between the groups with respect to the reasons named for sheep keeping: 89% of the households stated that the opportunity to acquire cash when needed was one of the reasons, but this reason was often cited together with keeping sheep for pleasure (28%) and for feasts. None of the households named generating a regular income as a reason. In M-3, livestock was used for savings, 68% of the households kept on average 9.7 head of cattle. Accordingly, saving money in a savings club was rare in M-3, while it was common in M-2, and practiced by nearly half of the households in M-1.

Flock productivity. Due to the high variation within groups no significant difference was found between the groups in seasonal and annual productivity (Table VI). However, the seasonal and the annual production index were higher in the flocks owned by M-2 households.
Table VI.

Flock Productivity Indices (FPI, g kg−1) by Socioeconomic Group ‘‘M’’ in Maroua (mean±SE)

 

M-1

M-2

M-3

 

FPI

(g kg−1)

n

(g kg−1)

 n

(g kg−1)

 n

Significance

Seasonal

128±30

13

240±53

10

194±45

10

n.s

Annual

816±132

25

999±315

12

739±153

19

n.s

Labor force of the households and labor sharing between members. The clusters differed in terms of available labor, time assigned to different activities, and labor sharing between household members, but due to the high variation and relatively small number of households with complete data the differences did not prove significant. The households in M-1 had a comparatively smaller labor force of 4.2 AME compared to 6.3 AME in group M-2 and 5.2 AME in group M-3. The labor invested in sheep per ewe equivalent in M-1 (134 h) was smaller than in the other two groups (290 h), and male and female household heads together spent almost the same amount of time looking after the sheep as did their children. Conversely, in M-2 most of the labor related to sheep was done by children younger than 15 years and the male household heads were not much involved. In M-3, a relatively high proportion of labor (64%) was performed by a paid herder, followed by male household members, while children under 15 years were only involved a little.

Differences and Similarities Between Sheep Keepers in Bobo-Dioulasso and Maroua

Maroua and Bobo differed in terms of the relative importance of agriculture: in Maroua 51% named either crop cultivation or livestock production as their main activity, while this was the case for only 14% in Bobo. Fifty percent of the households in Maroua cultivated land compared to 28% in Bobo. The size of sheep flocks was larger in Maroua (9.4 head in the baseline sample) than in Bobo (7.8 head). The proportion of traders among the interviewed households in Bobo was twice the proportion in Maroua. Retired officials represented 21% of the households in Bobo, but none were found in Maroua. In Maroua one-third of the households belonged to an ethnic group that traditionally kept livestock, while this was only the case for 14% in Bobo. Muslims dominated both samples; they represented 80% of the household heads in Bobo and 50% in Maroua. Household size was on average larger in Bobo (16 persons) than in Maroua (10 persons). There were fewer household heads with a modern school education in Maroua where it was highly probable that a person with modern school education would enter governmental service. Every governmental employee in the sample had a modern school education. Also more Christians than Muslims had a modern education; sheep keepers with modern education were often younger. These relations did not occur in Bobo.

Nevertheless, the most important factors characterizing the socioeconomic clusters, namely belonging to an ethnic group of traditional livestock keepers, status as a native or migrant, age, gender, and school education level of the household head were similar for both towns; the distinction between clusters was more clear-cut in Maroua. In both towns clusters formed by traditional livestock keepers were identified (M-3, B-3, and B-5), mainly headed by men with no modern school education, who kept larger herds of sheep and often cattle and also cultivated crops. Secondly, the groups of younger household heads with higher school education had the best knowledge of supporting organizations (M-2 and B-2); in Maroua they kept their sheep in fairly good conditions, and there was a trend towards higher productivity. The household heads of this group were recent migrants to Maroua, while those in Bobo were born in town. M-1 corresponded in its socioeconomic characteristics to B-1. The female headed households, of which the majority were assigned to M-1 and B-1, did not own cattle and their school education level was rather low. While three-quarters of the women in Bobo declared their principal activity to be housework, the choice of activities was more diverse in Maroua including livestock production, housework, crop cultivation, and trading.

DISCUSSION

In both towns cluster analysis revealed distinct socioeconomic groups of sheep keepers. The socioeconomic variables distinguishing the groups in Bobo were also valid for the sheep owners in Maroua. Although, the groups were not completely homogenous for the discriminating variables, this disadvantage of automatic multivariate analyses is outweighed by the advantage of enabling the unprejudiced inclusion of a multitude of socioeconomic variables, which is not possible in a purposive classification.

The hypothesis that socioeconomic groups of urban sheep keepers differed in their perception of urban sheep keeping was confirmed for Bobo. The differences related to the importance of sheep keeping, its future development in the town, and households’ plans regarding sheep keeping. In Maroua the groups differed mainly in their plans regarding sheep keeping and their perception of its continuity in their own family. The hypothesized differences in management practices and productivity between the groups were tested in Maroua and confirmed for management practices.

The results suggest that knowledge of socioeconomic characteristics does indeed help to identify livestock keepers with similar perceptions, husbandry practices and, thus, development potential. The respective clusters in Bobo and Maroua could be used for designing appropriate development interventions in these towns. In other cities similar studies could help policy makers and extension workers to determine appropriate target groups when planning interventions. Adequately addressing urban livestock keepers’ needs, however, is dependent upon their official recognition and legalization.

In both towns the heads of household that started sheep husbandry at an early age were born in town or were traditional livestock keepers. Contrary to the common belief that those farming in the city are the most recent in-migrants who have yet to integrate themselves into the urban economy, the longer-term urban residents were actually rather more likely to be engaged in farming (Ellis and Sumberg, 1998; Maxwell, 1995). This underlines the fact that motivation for livestock keeping is more complex than a mere household survival strategy, i.e., a response of the urban poor to limited alternative livelihood options and food insecurity. Sociocultural aspects have been mentioned by other authors, e.g., the religious importance of sheep for Muslim feasts in West Africa (Killanga et al., 2004). However, in Bobo and Maroua the pleasure of having animals as company was explicitly expressed as an important reason besides access to cash. Thus the benefits of sheep keeping should not be measured only by economic criteria.

The study also showed that besides vulnerable groups, such as female headed households, retired people, and people with limited formal education (Guendel and Richards, 2003), well-educated household heads were also involved in sheep keeping. These households were quite well-informed about existing services, although the knowledge of livestock husbandry was nearly always acquired within the family. Existing services were not tailored to the needs and circumstances of urban livestock keepers (Jahn et al., 1996), and especially not the urban poor (Guendel and Richards, 2003), who are often characterized by lower educational levels and greater time constraints (Ruel et al., 1999).

The more vulnerable households were more interested in improvements and exchange of information than the households with well-educated heads, but did not have free labor or capital resources to test new practices. Financial support or limited subsidies could help to overcome resource shortages if the required labor input was available. The ‘‘traditional livestock keepers’’ owning cattle and larger sheep flocks kept their animals in town in the same way as their rural ancestors did. They did not lack general livestock-related knowledge, but would need to be convinced to better adapt their practices to the specific urban conditions. The context of the town especially calls for consideration of social tensions arising as a result of problems associated with livestock keeping such as noise, odor, dirt, and hygienic risks.

CONCLUSION

In West African cities differences in motivation and management capacity between urban sheep keepers can be attributed to socioeconomic clusters. To assure accessibility and acceptance by a wide range of clients, extension agents offering livestock services should address these different target groups.

ACKNOWLEDGMENTS

This paper results from the collaboration between IRZV (Institut de Recherche Zootechnique et Vétérinaire) in Cameroon, CIRDES (Centre International de Recherche-Développement sur l’Élevage en zone Subhumide) in Burkina Faso and the Institute of Animal Breeding and Genetics, University of Göttingen in Germany. The study was part of the EU-funded project SECOVILLE (Socio-économie de l'élevage ovin péri-urbain). We are grateful to the above named institutions and especially to Mrs. Diara Kocty-Thiombiano from CIRDES for supporting this work.

Copyright information

© Springer Science+Business Media, Inc. 2006