Chapter 5: The study area: Socio-spatial organisation and land use

5.1 The household and the boma
5.2 Residence patterns
5.3 Water utilisation, grazing patterns and stocking rates
References

Two factors largely determine strategies for, and constraints on, livestock production in the study area: the group ranch to which the producer is affiliated and the wealth class of the household.

This chapter describes the socio-spatial organisation in the study site, including the household, the boma and the neighbourhood, and the interactions between residence patterns and resource utilisation.

The data presented were collected between 1980 and 1983, usually from sample households only, but sometimes from the whole population. Since household composition, livestock holdings, residence and herding patterns and the distribution of people and animals change over time, the numbers of livestock, households, bomas etc may not be consistent throughout. However, unless otherwise noted, the general patterns described pertain to the whole period under study.

5.1 The household and the boma

5.1.1 Household size and composition
5.1.2 Boma size and composition

5.1.1 Household size and composition

In the northern ranches (Olkarkar and Merueshi) there was a clear correlation between wealth (measured in terms of Tropical Livestock Units per active adult male equivalent - TLU/AAME; see Section 1.3.2: Producer heterogeneity and sampling design) and household size and composition. Rich households had 80% more people than poor households, a smaller percentage of whom were from the nuclear family, i.e. the producer, his wives and children (Table 5.1). The organisation of households is more fluid in Mbirikani, which made it more difficult to determine household size and composition¹. As a result, no clear relationship was found between wealth and household size. However, as in the northern ranches, rich households tended to have a smaller percentage of members coming from the nuclear family than did poor households.

¹ For example, it was common in Kisongo section (which includes Mbirikani) to find brothers who did not separate their families and animals after the death of their father, although each had his own inheritance. However, as each had the right to make decisions and to separate, they were defined as separate households.

Table 5.1. Average household size and proportion of nuclear members in household by wealth class and ranch.

Wealth class1	Olkarkar	Merueshi	Mbirikani
Poor	7.7 (80%)	9.0 (87%)	13.2 (82%)
Medium	9.9 (56%)	11.4 (66%)	10.2 (65%)
Rich	14.0 (62%)	16.2 (62%)	11.8 (63%)

Rich = ³ 13 Tropical Livestock Units (TLU) per active adult male equivalent AAME); medium = 5-12.99 TLU/AAME; poor = < 5 TLU/AAME.

Forty per cent of all households had resident mothers, step-mothers or siblings of the household head. Married sons remained with their fathers in 17% of all households, most commonly in richer households. A quarter of households had dependants who were not members of the nuclear family²; these are people who are incorporated into the household because they have insufficient resources to be self-supporting. In general, the wealthier the household the more dependants it had. Dependants represent a fairly broad spectrum of relationships to the household head. There were more dependants in Olkarkar than in other ranches, but the reason for this is not known.

² For the 24 dependency relationships for which information is available, the following is the distribution: six brother's families, four sister's families, four other agnatically related families, three returned married daughters, three mother's kin, one wife's kin, two other in-laws, one a father's friend (the dependency relationship was inherited).

Forty per cent of households borrowed children to help with herding or domestic tasks.³ The number of children borrowed did not differ between Olkarkar and Merueshi or among wealth classes, but poor households lent more children than did rich households (1.6 vs 0.4 children). Lending of children was not well recorded or! Mbirikani.

³ The arrangement may be a short-term emergency measure, but is more often a long-term one, with the child staying in the household of the borrower until marriage, in which case the borrower assumes the responsibility to feed, clothe, and help with required ceremonies (circumcision, marriage).

Maasai households traditionally joined with others, living together in a single boma, for various domestic and livestock management tasks, especially herding. There is an increasing trend toward individualisation in residence and production, especially among the Kaputiei in the north of the study area. In 1981 there were several single-household bomas on Olkarkar and Merueshi, but only one on Mbirikani (Table 5.2).

5.1.2 Boma size and composition

Bomas in Maasailand traditionally comprised 6 to 12 households (Jacobs, 1965; Njoka, 1979), but boma size has declined rapidly in the past 20 years⁴. By 1980, no boma in Merueshi had more than three households (Table 5.3), although 45% of households in Olkarkar and 60% of households in Mbirikani were in bomas of 4 or more households. In 1980 the mean number of households per boma was 2.7 on Olkarkar and 1.8 on Merueshi, which are in Kaputiei section, compared with 3.5 in Mbirikani, which is in Kisongo section. Between 1980 and mid-1983 the pressure for subdivision of Olkarkar ranch resulted in several bomas splitting (Grandin, 1987) and the mean number of households per boma on this ranch fell to 1.8. On Merueshi the number of households per boma fell slightly to 1.6 in mid-1983, while on Mbirikani boma size remained essentially unchanged (3.5 in 1980 and 3.6 in 1983).

⁴ Although decline has been a long-standing process, it is clear that in Kaputiei the biggest decrease in boma size and the emergence of single-household bomas came after the establishment of the group ranches. These phenomena seem related to the desire to stake a claim should subdivision of group ranches occur, and to a lesser extent, to increasing individualisation of production (see Section 3.3.7: Pressure for subdivision of group ranches). Group ranches in the Kisongo area are much newer; subdivision is not an issue in that area, and bomas on the whole are larger and thus more diverse.

Table 5.2. Residence types of sample households by wealth class and ranch.

Wealth class1	Residence type	Olkarkar	Merueshi	Mbirikani	Total
Poor	Alone	0	2	0	2
	With others	8	4	6	18
Medium	Alone	2	5	0	7
	With others	5	5	10	20
Rich	Alone	3	3	1	7
	With others	6	3	7	16
Total	Alone	5	10	1	16
	With others	19	12	23	54

¹ Rich = ³ 13 Tropical Livestock Units (TLU) per active adult male equivalent AAME); medium = 5-12.99 TLU/AAME; poor = < 5 TLU/AAME.

More households were sedentary in Kaputiei section than in Kisongo section. In 1981 more than 90% of Kaputiei household heads were living in their emparnat (the area where their fathers and grandfathers had lived), and the mean age of bomas was more than 3 years. In Kisongo, only 46% of household heads were living in their emparnat; the mean age of bomas was about 1 year.

In 1980 Maasai were still using a wide range of relationships to join bomas (Table 5.4). Producers in Olkarkar used a wider range of relationships than did those in the other ranches but close agnates tended to remain together when bomas subsequently divided, while less-closely related households left. In Merueshi, the trend to live with agnates was already well established. In Mbirikani, about half the households joined brothers, the other half joined friends.

As boma size declined in the north, so too did cooperation in herding and other routine management activities. This and other local implications of sedentarisation and individualisation of production are discussed in more detail in Section 5.2.2 (Neighbourhoods and reserved grazing areas) and Chapter 6 (Labour and livestock management).

Table 5.3. Distribution of boma size on Olkarkar, Merueshi and Mbirikani group ranches, 1980.

Households per boma	Per cent of households by boma size			Per cent of bomas by size category
	Olkarkar	Merueshi	Mbirikani	Olkarkar	Merueshi	Mbirikani
1	16	21	4	42	39	13
2-3	39	78	36	33	61	46
4 or more	45		60	25		40
Total	33	33	53	12	18	15

Table 5.4. Relationships used in joining bomas on Olkarkar, Merueshi and Mbirikani group ranches¹.

Relationship	Per cent of total recorded by ranch
	Olkarkar	Merueshi	Mbirikani
Clan	30	5	2
Close agnate2	17	64	51
In-law	17	11
Friend	17	21	47
Other	17

¹ For 1980; based on the single closest relationship to any other household in the boma.
² Father, brother, father's brother etc.

5.2 Residence patterns

5.2.1 Introduction
5.2.2 Neighbourhoods and reserved grazing areas

5.2.1 Introduction

Producers select a neighbourhood (and a boma) that best meets their goals, the needs of their animals and the preferences of their family. Herds need access to water and pastures, while families like to be near water, shops, schools and friends. The relative importance of these needs and desires in determining where to settle varies considerably with scale of production. For example, poor producers require less grazing than rich producers and are thus more likely to base their decision on where to settle on proximity to water and schools. They may, however, have to settle where there is a kinsman willing to help support them. In contrast, availability of grazing is of primary concern to richer households; finding sufficient grazing in a daily orbit is a qualitatively different problem for 500 cattle than for 50 cattle. The most important short-term considerations in choosing a place to live are:

· Proximity and freedom of access to water for human and animal consumption, the quality and the reliability of the supply and the labour necessary to extract and transport the water.

· Proximity to good grazing, the degree of competition from other livestock and wildlife, and the type of the terrain and fodder available between the boma and the water point.

· Availability of reserved grazing areas

· Proximity to schools and, occasionally, outlets for milk sales.

· Previous relations with potential neighbours.

Longer-term considerations differed between the north and the south of the study area because of differences in the mobility of households. Despite their high mobility, Mbirikani producers try to maintain a residence in their emparnat. In the north, the desire to stake a land claim and to choose a place one would like to settle permanently are more important considerations. In the north, for establishing a new boma, choice is largely circumscribed by the prior existence of other bomas and olopololis.

5.2.2 Neighbourhoods and reserved grazing areas

Neighbourhoods

Residential locations were close to permanent water sources on all three ranches (Figures 5.1 and 5.2). On Olkarkar, all five neighbourhoods were within 7 km of Simba Springs, leaving almost half the ranch without human settlement. On Merueshi, seven of the eight neighbourhoods were within 5 km of a water source: four were close to the pipeline on the western side of the ranch, three were in the north-east corner of the ranch and relied mainly on the shallow wells and boreholes associated with the Kiboko River. Of the nine neighbourhoods in Mbirikani six were close to the pipeline, while the other three depend on the Kikarankot River.

Neighbourhoods in Olkarkar comprised an average of three bomas, eight households, 86 people and almost 900 cattle and 850 smallstock. This is similar to the size of individual bomas in Maasai areas of Tanzania in the 1950s (Jacobs, 1965). Density of people and livestock varied in relation to proximity of the neighbourhood to amenities, grazing and water. Neighbourhoods also differed in wealth of households living there: for example, most poor households in Olkarkar were in neighbourhood 2 (Figure 5.1), which is close to water, shops and a school. Although the number of households per boma fell considerably in Olkarkar during the study period. This had relatively little effect on the population of neighbourhoods because most households stayed in the same neighbourhood. The distribution of bomas did, however, change, from closely clustered to more scattered as the new bomas established their own reserved calf pastures.

Figure 5.1. Neighbourhoods, reserved grazing areas and water sources on Olkarkar and Merueshi group ranches.

Figure 5.2. Neighbourhoods and traditional grazing management on Mbirikani Group Ranch.

Neighbourhoods in Merueshi were smaller than those in Olkarkar, with an average of roughly 60% as many households, people and stock (Tables 5.5 and 5.6). This was due in part to the greater dispersion of neighbourhoods in Merueshi but also in part to the greater desire for autonomous production and breakdown of traditional ties on this ranch. Mbirikani's neighbourhoods were much larger than those in the northern ranches, averaging nearly 8 bomas 21 households and 248 people (Table 5.7).

Proximity to water had a marked effect on the number of livestock per household in Olkarkar neighbourhoods. Households in the neighbourhood closest to water (neighbourhood 2) owned on average only 40 cattle, whereas those in the neighbourhoods farthest from water (neighbourhoods 3, 4 and 5) each owned some 150 cattle and 130 smallstock. Neighbourhoods in Merueshi were generally close to a water source and thus the effect of proximity to water on the number of livestock per household was less clear. There was no livestock census for Mbirikani as a whole.

Table 5.5. Characteristics of neighbourhoods on Olkarkar Group Ranch, 1980^a.

	Neighbourhood
	1	2	3	4	5	Total	Mean
Bomas	4	3	2	3	3	15	3.0
Households	11	11	4	3	10	39	7.8
People	136	91	65	48	88	428	86
Cattle	1553	413	673	1091	720	4450	890
Smallstock	1302	710	714	590	947	4263	853
Olopololis1	2	2	2	3	1	10	2

^a Excludes one boma with two households which is part of a neighbourhood in another group ranch.
¹ Olopololis are reserved calf pastures.

Table 5.6. Characteristics of neighbourhoods on Merueshi Group Ranch, 1980^a.

	Neighbourhood
	1	2	3	4	5	6	7	8	Total	Mean
Bomas	3	1	3	3	3	1	3	3	20	2.5
Households	5	1	5	5	6	2	4	4	32	4
People	53	52	65	55	52	9	78	63	427	53
Cattle	361	654	802	752	471	216	864	120	4240	530
Smallstock	498	319	652	654	546	50	534	410	3663	458
Olopololis	3	1	1	2	1	1	1	2	12	1.5

^a Excludes four households that took up residence in adjacent ranches in late 1980.

Table 5.7. Characteristics of neighbourhoods on Mbirikani Group Ranch, 1980.

	Neighbourhood
	1	2	3	4	5	6	7	8	9	Total	Mean
Bomas	9	5	15	12	8	10	3	4	3	70	7.8
Households	25	12	49	32	24	17	6	13	8	186	20.7
People	300	144	588	384	288	204	72	156	96	2232	248
Olopololis	3	1	2	2	2	-	1	2	2	15	1.7

Note: An average household in Mbirikani has 12 people. Livestock data for the entire group ranch were never collected, because of the size of the area and the mobility of its stock and people.

Reserved grazing areas

The Maasai have long set aside pastures near residential areas for the exclusive use of calves and weak animals. These areas of reserved grazing are known as olopololis. Establishment of olopololis is controlled by the council of elders in each neighbourhood.

In 1982 there were 13 olopololis in Olkarkar, with an average area of 162 ha and covering 20% of the ranch (Table 5.8). Ten of them were each used by a single boma, seven of them each by a single household. However, three olopololis retained the attributes of a neighbourhood olopololi; one was used by 11 households in six bomas the second by seven households in three bomas and the third by 10 households in three bomas. The last lay partly inside Kiboko group ranch and was used also by a household in that ranch.

The 13 olopololis in Merueshi had an average size of 350 ha and accounted for 25% of the ranch(Table 5.8). Nine of them are each used by only one boma, while none was used by more than three bomas Seven of the olopololis were each used by only a single household. Each olopololi was used by an average of two households; none was used by more than four households.

Table 5.8. Characteristics of reserved grazing areas (olopololis) on Olkarkar Merueshi and Mbirikani group ranches, 1982.

		Ranch
		Olkarkar	Merueshi	Mbirikani
Number		13	13	15a
Area
	mean (ha)	162	350	570
	range (ha)	47-403	155-800	200-1600
	% of ranch area	20	25	5

^a Includes one olopololi servicing a primary school.

Mbirikani group ranch had 15 olopololis covering about 5% of the ranch (Table 5.8). The olopololis were large, averaging 570 ha, and were each used by an average of four bomas and 11 households. Two were each used by only one boma, but none was used by a single household.

The changes in the use and management of olopololis in the study area are demonstrated by those occurring in Olkarkar between 1979 and 1983 (Grandin, 1987). In 1979 Olkarkar had nine olopololis, only one of which was controlled by a single household (Table 5.9). Four were controlled by residents of a single boma, comprising a total of 12 households. The remaining four were shared by more than one boma, and approximated neighbourhood control.

By 1983, the number of olopololis had increased to 15. Most of the increase was in single-household olopololis. Although there were still four single-boma olopololis, the bomas each comprised only two households headed by full brothers. Three of the olopololis shared by more than one boma were shared by many households and could still be classified as neighbourhood olopololis. The fourth was now shared by two bomas formed when two brothers had separated after the death of their father, each establishing his own boma but sharing their father's olopololi.

Table 5.9. Changes in the number of olopololis and their use on Olkarkar Group Ranch, 1979-83.

	No.	Users
		Per cent of ranch area	Single household	Multiple household
				1 boma	> 1 boma
1979	9	13a	1	4	4
1983	15	20	7	4b	4c

^a Estimated.
^b All of these consist of two full brothers only.
^c Three of these are neighbourhood bomas while the other includes two bomas of two brothers and two of their sons.

The proliferation of olopololis in the northern ranches was related more to their use in establishing rights over land than to their value as a management tool. Hence, the size of the olopololis bears no necessary relationship to the needs of the "owning" household or households.

In conclusion, between 1979 and 1983 there was a proliferation of single-household olopololis in the northern ranches. This has implications for livestock management, in particular because many producers are using their olopololi to feed stock other than calves.

5.3 Water utilisation, grazing patterns and stocking rates

5.3.1 Water utilisation in the northern ranches
5.3.2 Grazing patterns and stocking rates in the northern ranches
5.3.3 Grazing patterns and stocking rates in the southern ranch

The distribution of water points in the study area was outlined in Chapter 4 (see Section 4.5: Water resources). This section discusses the use of these sources in Olkarkar and Merueshi, and describes the use of different water sources and the patterns of livestock movement in Mbirikani.

5.3.1 Water utilisation in the northern ranches

Simba Springs is the only permanent water source in Olkarkar and 79% of all visits to water points were to the Springs (Table 5.10). In contrast, there are several permanent sources of water in Merueshi, resulting in more varied patterns of use. Neighbourhoods in the north-west (1,6 and 7; see Figure 5.1) went mostly to the pipeline (60% of visits). Neighbourhoods in the north-east exploited the shallow wells in the Kiboko riverbed (60% of visits). Neighbourhood 8 used the nearby borehole. The single household in neighbourhood 2 used both the pipeline and the shallow wells. Seasonal sources were used mostly in the rainy seasons and were more important in Olkarkar than in Merueshi, where ponds were used in the western and central portion, and river pools were used in the south. In a normal year these sources accounted for 30% of total use by the neighbourhoods in their vicinity.

Aerial surveys in the dry periods in February and June 1982 showed that more than half the cattle and three quarters of the small ruminants on the two ranches were within 5 km of one of the permanent sources of water (King et al, 1985). Concentrations of stock in the eastern parts of the ranches were higher in June than in February. There were considerable eastward movements of stock within Merueshi, and there was an influx of livestock from Mbuko ranch. Several herds from Mbirikani and Kimana ranches grazed in Merueshi and the adjacent Chyulu foothills in the north-east.

Table 5.10. Utilisation of watering sources on§§Olkarkar and Merueshi group ranches, June 1981 to April 1983.

		Per cent of all visits to water points
		Olkarkar	Merueshi
Permanent
	Simba Springs	79
	Pipeline	2	30
	Boreholes		16
	Wells	1	29
Seasonal
	Ponds	6	20
	Rivers	12	5

5.3.2 Grazing patterns and stocking rates in the northern ranches

Daily movements to grazing of herds belonging to sample households in the two northern group ranches were recorded every 2 weeks between July 1981 and June 1983. Grazing pressure was based on the total livestock population resident in each ranch in mid-1982. It was assumed that unsampled households within each cluster were practising the same grazing management and movement patterns as their sampled neighbours. Thus, for each cluster the ratios between total stock and sampled stock were calculated, separately for cattle and smallstock. These ratios were derived from the initial survey in 1980-81, which included livestock populations of both sampled and unsampled households (ILCA, 1981).

Based on herd and flock structure data from King et al (1984), total cattle of each household were subdivided in 65% adult castle, 25% weaners and 10% suckling calves. Similarly, it was assumed that grazing flocks comprised 80% of the total, the remainder being lambs and kids. As suckling stock were kept around the boma and did little grazing, they were excluded from the analysis (see Semenye, 1987; de Souza and de Leeuw, 1984).

Weights were assigned to each class: 250 kg for adult cattle, 120 kg for immatures and 25 kg for smallstock. From these weights, total grazing mass of livestock in each ranch was calculated. Frequencies of visits by each household to grazing areas were multiplied, first with the appropriate stock number by class for each and then by the ratio between sampled and total households by cluster. These weighted frequencies produced the grazing pressure by location and by aggregating grazing locations for each zone.

Grazed livestock in Tables 5.11 and 5.12 refers only to the resident livestock within each ranch territory; herds grazing in other ranches or immigrant herds have not been included in the calculations of grazing pressure. There is, however, considerable grazing across the boundaries into Poka and Kiboko ranches and ranch territories have been enlarged somewhat to allow for this movement (Figure 5.3).

Grazing locations within each ranch were aggregated into six grazing zones in Olkarkar and four grazing zones in Merueshi (Figure 5.3). As would be expected, the distribution of these zones was similar to that of neighbourhood clusters (see Figure 5.1).

Table 5.11. Grazing pressure by grazing zone on Olkarkar Group Ranch.

		Zone						Total
		I	II	III	IV	V	VI
Area (% of ranch)		15	12	11	18	15	29	100
Livestock (% of total TLU)
	owned	34	11	15	40			100
	grazed	30	12	15	12	22	9	100
Grazing pressure								Mean
	kg/ha	234	119	161	90	174	34	119
	ha/TLU	1.1	2.1	1.6	3.2	1.4	7.4	3.5

Table 5.12. Grazing pressure by grazing zone on Merueshi Group Ranch.

		Zone				Total
		I	II	III	IV
Area (% of ranch)		22	29	38	11	100
Livestock (% of total TLU)
	owned	24	48	28		100
	grazed	32	50	16	2	100
Grazing pressure						Mean
	kg/ha	100	90	21	7	57
	ha/TLU	2.5	2.8	12.1	35.2	9.8

Figure 5.3. Stocking rates on Olkarkar and Merueshi, 1981-83.

On Olkarkar, stocking rate declined radially away from Simba Springs. Within the northern part of the ranch livestock biomass was fairly evenly distributed, although grazing pressure was highest in zone I and zone V, the main grazing areas for the richer households in neighbourhoods 1, 4 and 5 (Figure 5.1). These two zones accounted for 30% of the ranch and more than half its total livestock biomass (Table 5.11). The five zones in which neighbourhoods were located (zones I to V) accounted for 71% of the ranch and had an average stocking rate of 1.6 ha/TLU. Zone VI was less used because it is far from both Springs and the pipeline and because its vegetation consists largely of coarse tall grasses.

The utilisation of grazing resources in Merueshi was different from that in Olkarkar, because bomas were mainly located along the ranch periphery and reserved grazing areas were more evenly distributed (Figure 5.1). There was high grazing pressure in zones I and II which cover 50% of the ranch but accommodated 82% of all stock; this converts into an overall stocking rate 2.7 ha/TLU (Table 5.12, Figure 5.3). This high pressure was in contrast to the low grazing use in zone lilt Although the five households resident in this zone owned 28% of the ranch livestock, they herded their animals within zone III itself for only half the study period. There are several interconnected reasons for this mobility. About 80% of the cattle in zone III were owned by one household (1100 head in 1982) and this herd would overgraze the zone if it grazed there permanently. Grazing pressure in zone IV was low because there were no settlements there, it was relatively far from water and was regarded as a fall-back area during dry periods. It was heavily grazed during the 1983-84 drought (Grandin et al, 1989).

From this analysis, it is evident that Merueshi was much more lightly stocked than Olkarkar; about 5 ha/TLU as compared with 2 ha in Olkarkar. Although historic reasons may have played a part, it is argued that this difference in the overall utilisation rates reflected the differences in grazing resources between the two ranches. On average the plant cover in Olkarkar is much denser than in Merueshi. On Olkarkar most of the land consists of undulating uplands over volcanic rock, which supports a relatively dense cover with desirable grasses, some of which are resistant to repeated grazing (see Figure 4.2 and Table 4.2). Only a small part, mainly in the east, has soils over basement complex, on which much more open grass communities are found. In Merueshi the more productive rangelands cover less than half the ranch and are concentrated mainly in the north and the east. This good cover contrasts with the sparse vegetation in the SW portion of the ranch (see Section 4.2: Landscapes, soils and vegetation). This resource gradient running approximately from the north-east to the south-west is reinforced by the rainfall gradient along the same direction (see Section 4.3: Climate).

5.3.3 Grazing patterns and stocking rates in the southern ranch

Traditionally, the Kisongo Maasai have divided their land into well-defined residential and grazing areas. The residential areas and the permanent bomas are usually as close as possible to permanent water and about half the ranch area was designated residential land; it also contained the neighbourhoods, all olopololis and stretched 510 km in width on either side of the pipeline and the Kikarankot River with its associated swamps

Figure 5.2 shows the distribution of the different grazing areas. The arrows denote the sequence in which areas were used through the dry season. The source of the arrow is the neighbourhood and the head of the arrow marks areas for grazing in the late dry season. At a distance of 5-10 km from the neighbourhood sites there were areas earmarked for grazing in the early dry season, while further away there was a belt for use later in the dry season. At the margins of the early dry-season zone temporary camps were often constructed 10 to 15 km away from the pipeline and herds were put on a 2-day watering regime.

While grazing rights and use are well recognised for the residential areas and their olopololis, user rights became more fluid with increased distance. The bomas that were associated with these areas of deferred grazing did not have exclusive usufruct rights but they collectively decide when livestock may enter an area for grazing. In times of good rainfall these final dry season areas would not be entered before the next rains fell. Ideally, rains would be sufficient to fill surface pools in the most distant wet season areas, allowing cattle to proceed there, and thus preserve the grazing in the residential areas and in the olopololis. This grazing system was in operation when most herds were resident within the boundaries of the ranch and rainfall was normal.

However, this traditional system described above has been disturbed in the western part of the ranch primarily because of the construction of Risa water tank just outside the western boundary of the ranch. This area is located west of the pipeline and stretches across the somewhat arbitrary western boundary and the traditional dry season areas of the western part of Mbirikani ranch. Before the construction of the water tank, cattle moved westwards and southwards from the residential areas along the pipeline. As the dry season progressed, herds would go closer to the seasonal Kiboko River, eventually crossing it and grazing west of it. When the rains came cattle would water either at the river or further west at one of the many water pools north of Amboseli Park. After the Risa water tank was built, permanent bomas were constructed nearby and the new occupants of this area developed a north-eastward pattern of grazing. Their cattle thus met and competed for grazing with cattle moving westwards from the neighbourhoods along the pipeline. If the temporary waterholes north of Amboseli were full and allowed grazing to continue into the dry season, then the area north-east of the Risa water tank was not under severe pressure. If the rains are poor, herds moved outwards from the permanent sources of water (pipeline and Risa tank) early in the dry season, which led to early competition between the two opposing movements of cattle.

Figure 5.4. Distribution of cattle herds on Mbirikani in February 1982

Although rainfall during 1981 was somewhat below average, it may be considered a fairly typical year. In April 1981 the rains caused the formation of surface water pools in many parts of the ranch so that grazing was possible close to the Chyulu Hills (Figure 5.2). The livestock distribution showed little change from June to August and remained stable until the end of the dry season. The first rains in November and December 1981 were low and localised which caused the clumping of herds and flocks in several areas, a situation that continued to February 1982 (Figure 5.4). Thus, throughout most of 1981 stock relied on the northern stretch of the pipeline (between Makutano and Olandi), and the swamps along the Kikarankot River and the boreholes along the Kiboko River. Several herds grazed in Mbuko and Merueshi territory, while during June to August 1981 a few herds used the Risa tank close to the Amboseli National Park. In November 1981 there was a sudden move from rangelands east of the pipeline to the western boundary of the ranch because good rainfall had filled the shallow waterholes there. Thus, during 1981 75% of the sample cattle herds remained within the ranch territory.

Smallstock were managed differently from cattle in that they stayed mostly within 5 km of the pipeline. Three flocks joined the cattle herds around Risa tank. Like cattle smallstock made little use of areas in the south-west, except for a few flocks which went first to a tributary of the Kikarankot river in August 1981 and then moved to the Acacia tortilis woodland east of Kimana again relying on pipeline water.

While during 1981 most livestock remained within the boundaries of the ranch, the low rainfall in late 1981 and the even poorer rains in early 1982 caused wholesale shifts of the livestock population to grazing land outside the ranch, both towards the south and to the north. Patterns of herd movement and the population estimates were derived from aerial surveys for three distinct periods in 1982 (King et al, 1985). As was done for the northern ranches, Mbirikani ranch was subdivided into grazing zones that follow as closely as possible the traditional grazing areas: zones 1, II and III represent the residential areas whereas the other zones (IV to VIII) coincide with the dry season grazing areas to the east and the west of the pipeline (Figure 5.5).

Even though in February 1982 these movements had already started, dispersal within the ranch still corresponded to the dry season distribution shown in Figure 5.4. Over half the cattle were still relying on the pipeline but use of its southern section was much greater than in the previous year. From February onwards the exodus got underway properly. Most herds went first to the swamps, either those near the southern pipeline section or to the Chyulu foothills relying on the water points in the eastern swamps using a 2-day watering regime; about 20% (of the 42000 head estimated during the aerial survey) followed the latter strategy. As a result of the exhaustion of the fodder supplies surrounding the swamp zone, herds moved further to the southwest and by mid-June 57% were grazing in Kuku Ranch using either the remaining water pools along the Loolturesh river or the wells near Iltilal (14%). Towards August 1982, these pools were drying out and the reliance was shifted to the wells.

Smallstock followed an itinerary similar to cattle except that they moved gradually southwards along the pipeline and then moved straight into Kuku Ranch and the Iltilal well zone without stopping in the swamp zone. As a result, the western and central parts of Mbirikani were almost entirely evacuated. Only 11000 cattle and 1300 smallstock remained along the northern pipeline and its adjacent grazing area in the north-east. As some 140 households have their permanent bomas in zone I, it was calculated that about eight cattle per household remained behind. These represented mainly lactating cattle and their calves to feed the resident family members. In the residual areas around the swamps (zone III) another 6000 cattle and 3000 smallstock remained.

Good rainfall in late October and November 1982 not only produced abundant new forage (see Table 4.5), but also filled most of the ephemeral ponds and riverbeds on the ranch, encouraging the return of herds and flocks. By late December, all but 7% of the livestock population had come back within the ranch, but some had not reached their permanent bomas along the pipeline. Nevertheless, 70% of all cattle and 65% of all smallstock were counted within the three residential zones and in zone I cattle had already reached a density of close to 50 TLU/km² or 2 ha/TLU (Figure 5.5). The remainder was dispersed over the dry season areas in particular in the areas to the west; this is in contrast to the distribution in February 1982 when grazing pressure was high in the east. The eastern area (zones IV and V) accounted for only 11 % of the cattle and 8% of the smallstock.

The effect of these stock migrations on the overall stocking rates of the ranch is shown in Table 5. 13. While in February and December cattle numbers were similar indicating that by December 1982 most herds had returned, in June only 40% of the cattle and less than 30% of the smallstock remained on the ranch. This proportion was even lower between June and November (Peacock, 1984). During February and December the average stocking rate of domestic herbivores was between 5.1 and 5.4 ha/TLU, while in June the rate dropped to 12.7 ha/TLU. The distribution of herbivores over the grazing zones showed that in February high stocking rates occurred along the southern end of the pipeline (zone 11) and in the residential areas North of the swamps and rivers (zone III). Grazing pressure was also high in the north-east (zone IV) indicating that many herds were on a 2-day watering regime (Figure 5.5).

It appears that grazing strategy of maximum dispersal and the resultant distribution was much influenced by the influx of wildebeest and zebra at the start of the rains. While in February and June these two species accounted for respectively 2 and 8% of the total herbivore biomass, this proportion rose to 18% in December 1982 (Table 5.13). More importantly, over 80% of all wildlife were found in the residential areas along the pipeline and its adjacent dry season area in the east. In zones IV and V, 42% of the total herbivore biomass consisted of wildlife and they competed heavily for the available forage resources and were instrumental in keeping away cattle from the eastern dry-season zones.

Figure 5.5. Stocking rates on Mbirikani in February 1982.

Figure 5.5. Stocking rates on Mbirikani in June 1982.

Figure 5.5. Stocking rates on Mbirikani in December 1982.

Table 5.13. Seasonal herbivore populations and stocking rate on Mbirikani Group Ranch, 1982.

This account shows that during good rainfall seasons and their aftermath, Mbirikani herds and flocks stayed within the ranch resulting in stocking rates in residential areas that are well beyond the carrying capacity. This necessitated rigorous grazing control that encouraged dispersal of stock towards less heavily utilised areas. Concomitantly, it requires the adoption of 2-day watering regimes. It is also clear that swift movements to grazing lands with ephemeral water ponds whenever they fill is an essential part of the same strategy, as it further assists in alleviating the grazing pressure in the areas closer to permanent water.

References

Grandin B E. 1987. Land tenure, sub-division and residential changes on a Maasai Group Ranch. Development Anthropology Bulletin 4(2):9-13.

Grandin B E, de Leeuw P N and Lembuya P 1989. Drought, resource distribution and mobility in two Maasai group ranches in southeastern Kajiado District. In: Downing T E, Gitu K W and Kamau C M (eds), Coping with drought in Kenya. National and local strategies. Lynne Rienner Publishers, Boulder, Colorado, USA. pp. 245-263.

ILCA (International Livestock Centre for Africa). 1981. Introduction to East African range livestock systems study/Kenya. ILCA/Kenya Working Document 23. ILCA, Addis Ababa, Ethiopia. 71 pp. [ILCA Library accession no. 31496]

Jacobs A. 1965. The traditional political organization of the pastoral Maasai. PhD thesis, Nuffield College, Oxford University, Oxford, UK. 417 pp.

King J M, Sayers A R. Peacock C P and Kontrohr E. 1984. Maasai herd and flock structure in relation to livestock wealth, climate and development. Agricultural Systems 13:21 -56.

King J M, Sayers A R. Chara P. de Leeuw P N. Peacock C P. Fillo F and Maehl J H H. 1985. Improving aerial counts of Maasai livestock. Agricultural Systems 16:231-256.

Njoka T J. 1979. Ecological and socio-cultural trends of Kaputiei group ranches in Kenya. PhD thesis, University of California, Berkeley, California, USA. 318 pp.

Peacock C P. 1984. The productivity of small stock in three group ranches in Kajiado district, Kenya. PhD thesis, Reading University, Reading, UK. 384 pp.

Semenye P P. 1987. Factors influencing Maasai cattle productivity and nutrition in Kajiado district, Kenya. PhD thesis, University of Nairobi, Nairobi, Kenya. 326 pp.

de Souza A P M and de Leeuw P N. 1984. Small stock use of reserved grazing areas on Merueshi group ranch. In: Proceedings of the Third Small Ruminant CRSP Workshop. Small Ruminant Collaborative Research Support Program, Nairobi, Kenya. pp. 289-299.