10.1 Inputs for the simulation models
10.2 Results
References
In extensive rangeland systems, livestock production is highly dependent on the availability of natural grazing, the quantity and quality of which are primarily determined by the amount and distribution of rainfall, given the temperature regime, soil-type and topography of a particular rangeland site. In eastern Africa, rainfall fluctuates widely from year to year.
The results reported in the preceding chapters were recorded mostly during a 2-year period following a succession of years in which rainfall was relatively favourable to primary production. However, over the past 100 years severe droughts have occurred at least once in every 8-12 years. This causes enormous fluctuations in the productivity of pastoral systems. Thus short-term studies, such as that conducted by ILCA in Maasailand from 1981 to 1984, cannot provide a complete picture of the dynamics of pastoral livestock production. This chapter attempts to examine the long-term variation of the Maasai livestock production system by using forage and livestock production models.
The strong linkage between herd productivity and the quality and quantity of the fodder supply has been commented upon throughout this study. What is less easy to establish is the range of variation for each cattle productivity parameter, particularly calving rate and mortality. These parameters have been predicted with biological herd simulation models for several pastoral production systems (Sullivan et al, 1981; de Leeuw and Konandreas, 1982). However, it is difficult to apply such biological models to pastoral systems (see Wagenaar and Kontrohr, 1986; de Leeuw, 1986). Stochastic models have also been used to predict primary productivity of rangelands using probabilities of annual rainfall distributions. However, linking such a stochastic model with a biological livestock production model was considered too complex and impractical.
The approach taken here was to use actual climatic data to estimate lengths of growing seasons. Forage production was estimated from these lengths of growing seasons. Estimates of cattle productivity were then based on these estimates of forage production.
Herd projection models were developed for the three wealth classes of producers on a 10000 ha group ranch using the data for Olkarkar. The models were applied to herds of 30, 60 and 300 head of cattle, representing the mean holdings of poor, medium-wealth and rich producers. The models generated changes in herd size, stock losses and saleable stock and simulated annual and long-term livestock and milk offtake for these three herd sizes; they also identified changes in these parameters according to year type.
The results of the herd models were then aggregated to arrive at the output for the entire Olkarkar Group Ranch by weighting them in accordance with the frequency distribution of these herd sizes in the ranch. Two assumptions were made for aggregating the output in this fashion. The first was that the 30 years for which the future projections were made (1983-2012) would have a similar pattern of growing seasons as that observed between 1957 and 1986. The second assumption was that the proportions of poor, medium-wealth and rich producers on the ranch would remain the same as those observed during the 1981 -83 period, which will of course not be the case as households will change as household heads grow old and sons divide the herd.
10.1.1 Fodder resources
10.1.2 The herd-projection model
10.1.3 Long-term milk supplies
10.1.4 Culling, sales and purchase policies
Growing-season duration was calculated using a soil moisture balance model developed by Musembi (1984; 1986). This model is similar to that used by Potter (1985). Estimation of herbage production in relation to the length of the growing season was based on an analysis of data from several range areas in semi- arid eastern Africa (de Leeuw and Nyambaka, 1988).
There are two marked growing seasons in eastern Kajiado, the first rains from October to January and the second rains from March to May. There is a dry season of variable length between these two rainy seasons, and the second rains are followed by a long dry season lasting from June to early October. in the short term, grazing resources are determined by the combined durations of the two growing seasons, while longer-term trends depend on the variability of annual growing period over longer time-spans.
Growing-season durations were calculated from data covering a 50-year period (1935-84) from two rainfall stations (Makindu and Simba) representative of the eastern portion of Kajiado District. The frequency distributions of the length of the two seasons were markedly different. For the first season, growing periods of 2 months or more occurred in 44% of the 50 years, while short seasons of one month or less prevailed in another 28% of the years (Figure 10.1). The mean over the 50-year period was 1.7 months. For the second rains the proportion of short seasons was much greater: in 54% of the years the growing season lasted 1 month or less whereas seasons of 2 months or more occurred only in 1 year in 3 (Figure 10.1). The mean duration of the second rainy season was 1.2 months.
Roughly 1 year in 3 had an annual growing period of 2 months or less, whereas 1 year in 4 was wet with at least a 4-month growing season (Figure 10.2). The mean annual growing period was 2.9 months for the whole 50-year period.
Using year-types as single events to predict resource conditions ignores carry-over effects from previous years. A very dry year after a series of wet years would have much less effect on livestock productivity than if the same dry year followed several years of below-average rainfall. Year-types as defined by the length of the annual growing season were plotted for a 30-year period (Figure 10.3).
Herbage yields per annum were estimated using durations of the total annual growing season as predictors (Table 10.1) (Potter 1985; de Leeuw and Nyambaka, 1988). Production was 1.5 t DM/annum or less in about a third of the years and 3.0 t DM/ha or more in about a third of the years (Figure 10.4).
This section discusses the various inputs used in this model, together with the assumptions for culling, sales and livestock purchasing policies.
Herd composition
The initial herd composition specified at the start of the model was derived from the data for Olkarkar Group Ranch (King et al, 1984). The composition of the two smaller herds was similar, while that of the herd with 300 head had a smaller proportion of young females and adult cows and three times as many steers more than 3 years old (Table 10.2).
Calving percentage
Breeding females were defined as all adult cows and a varying proportion of 3- to 4-year-old heifers.
Table 10.1. Expected daily herbage growth rates and seasonal productivity for rangelands in eastern Kajiado.
|
Duration of growing season |
Growth rate |
Seasonal yield |
|
0.5 |
13 |
0.2 |
|
1.0 |
17 |
0.5 |
|
1.5 |
22 |
1.0 |
|
2.0 |
25 |
1.5 |
|
2.5 |
30 |
2.3 |
|
3.0 |
30 |
2.7 |
|
3.5 |
28 |
2.9 |
|
4.0 |
27 |
3.1 |
In drier years none of these heifers conceive, whereas in good years 10-20% of them do. The calving percentage is governed mainly by year-type. During dry years, conception rates are low, causing a small calf crop in the next year, while high calving percentages mostly prevail immediately after drought because many of the surviving cows are open and likely to conceive once forage conditions improve. Overall mean calving rate was 51%.
Mortality
Mortality rates were specified for each animal class for each year, assuming that mortality rate is primarily determined by feed availability rather than disease incidence.
Table 10.2. Initial composition of herds comprising 30, 60 and 300 head.
|
|
Herd size (no. of animals) |
|||
|
30 |
60 |
300 |
||
|
Herd composition (% of herd) |
||||
|
Males |
|
|
|
|
|
Calves 0-1 year |
9 |
8 |
8 |
|
|
Steers 1-2 years |
8 |
10 |
7 |
|
|
|
2-3 years |
8 |
6 |
9 |
|
|
3-4 years |
3 |
3 |
9 |
|
|
> 4 years |
1 |
2 |
|
|
Breeding bulls |
3 |
5 |
3 |
|
|
Total males |
31 |
33 |
38 |
|
|
Females |
|
|
|
|
|
Calves 0-1 year |
9 |
9 |
9 |
|
|
Heifers 1-2 years |
8 |
11 |
8 |
|
|
|
2-3 years |
8 |
10 |
8 |
|
|
3-4 years |
11 |
8 |
8 |
|
Adult cows |
33 |
29 |
29 |
|
|
Total females |
69 |
67 |
62 |
|
See Tables 7.1 and 7.2 for comparison.
The 30-year mean, minimum and maximum mortality rates for each of the 10 stock classes are shown in Table 10.3. Minimum rates were applied during favourable periods whereas the peak rates were applied during drought periods. Heifers and steers had mortality rates ranging from 4% to 30%. In most years death rates were below 10%, and in four of the years between 10% and 20%. The range of mortality rate in cows was much larger than in growing stock over 1 year old. In 7 out of 10 years less than 10% died, but in drier years the death rate was 11-20%, reaching 40% in drought years. Calves had a minimum mortality of 10% in half the years and higher rates in the other half, up to a maximum of 60% during drought.
Table 10.3. Mean, minimum and maximum mortality and liveweights by age/sex class.
|
|
Mortality |
Weight (kg) |
||||
|
Mean |
Min |
Max |
Mean |
Min |
Max |
|
|
Cows |
9 |
4 |
40 |
266 |
230 |
300 |
|
Calves |
15 |
10 |
60 |
59 |
40 |
75 |
|
Heifers 1-2 years |
9 |
4 |
30 |
135 |
100 |
150 |
|
Heifers 2-3 years |
8 |
4 |
25 |
180 |
130 |
210 |
|
Heifers 3-4 years |
7 |
4 |
20 |
220 |
170 |
260 |
|
Steers 1-2 years |
8 |
4 |
30 |
145 |
110 |
170 |
|
Steers 2-3 years |
8 |
4 |
25 |
200 |
150 |
230 |
|
Steers 3-4 years |
7 |
4 |
20 |
250 |
200 |
290 |
|
Steers >4 years |
6 |
4 |
20 |
340 |
300 |
380 |
|
Breeding bulls |
6 |
4 |
15 |
340 |
300 |
380 |
Weight changes
Mid-year weights of all age/sex classes in the simulated herds were required for each of the 30 years to calculate herd biomass production and aggregate grazing pressure. These weights were derived from King et al (1984), who weighed some 5000 came in all three group ranches in 1980-81. Minimum and maximum weights were indicative of those that would occur in very dry and very wet years (Table 10.3). These weight changes were taken into account in calculations concerning the balance between grazing resources and their utilisation by herbivores (see Section 10.2.1: Herd size and stocking rate).
The model estimated the potential availability of milk in relation to year-types. The factors that affect the actual milk supplies for household subsistence were discussed in Chapter 7 (Section 7.1.7: Milk offtake and lactation yield) and Chapter 8 (Section 8.4: Milk, food consumption and nutritional status). Milk supply depends foremost on herd size and in particular on the potential number of lactating cows, i.e. cows with a calf at foot. The number of lactating cows was generated by the herd-projection models, based on the number of calves in the herd in the middle of each year. The reduction of milk yield due to calf and cow mortality was thus accounted for by apportioning the mortality equally over the first and the second halves of the year.
The annual potential milked-out yield per cow was derived from monthly milk offtake data with adjustments for the number of cows milked and milking frequency (see Section 7.1.7: Milk offtake and lactation yield). Subsequently, monthly offtakes were aggregated for each rainy season and for each year for the entire 30-year period.
Milk-offtake profiles per cow by month are illustrated in Figure 10.5 for six selected year-types, ranging from very dry to wet. Bars represent average monthly yield per cow taking into account the fact that in dry months some cows are not milked at all or are milked less than twice a day. Potential milk production for each month varies with the length of each growing season and thus by year-type. Years with short growing seasons, totalling less than 2 months, have short periods with reasonable offtake and up to 5 months with no milk at all (Figures 10.5a and 10.5b). When the total annual growing period was between 2 and 3 months long, monthly milk yields exceeded 15 litres per cow for 6 months (Figures 10.5c and 10.5d), whereas in good years (annual growing period of more than 4 months) yields exceeded 20 litres per cow per month throughout the year (Figures 10.5e and 10.5f).
Annual milk yield per lactating cow ranged from about 60 litres in the worst year to 360 litres in the best year.
To summarise the impact of year type on the herd productivity parameters, year-types were grouped in four forage resource classes (Table 10.4). Three of the 30 years were classed as very low, 12 as low, 10 as medium and 5 as high. Over this range, annual rainfall rose from 307 mm to 830 mm, with a mean of 550 mm, and the annual growing period increased from 1 month to almost 5 months.
The mean values of the came productivity parameters that were used in the projection model are given in Table 10.5 for each of the forage resource classes. The largest differences between resource classes were in annual milk yield and mortality rates. Average calving percentage in a given year was less influenced by forage resources during that year because of the time-lag between conception and parturition.
The Maasai cull cows when they are 8 to 12 years old. For the model a policy of culling and selling 10% of the cows yearly was adopted. Breeding bulls were culled at a faster rate of 25% per year to avoid in-breeding. Since sales policies materially affect the long-term productivity of a given herd, it was decided to hold constant the total number of animals sold across years in order to minimise the effects of differential sales policies. The actual mean numbers of animals sold as observed during the 1981-83 study (4, 7 and 17 head per year for poor, medium-wealth and rich producers, respectively) were initially used in the model. A sensitivity analysis of different sales strategies was conducted on the 60- and 300-head herd models, and this is discussed in Section 10.5 (Effects of increased offtake of steers on herd and ranch productivity). The types of animal sold was determined by a decision rule that first sold all the cull cows and bulls. If there were fewer of these than the fixed number required for sale the difference was made up by selling steers of 4 years old or older or, if there were too few of these, younger steers.
Table 10.4. Rainfall, length of growing season and forage yield for year-types grouped by resource classes.
|
|
Resource class1 |
Mean |
|||
|
Very low |
Low |
Medium |
High |
||
|
Rainfall (mm) |
|
|
|
|
|
|
1st season |
178 |
221 |
431 |
550 |
340 |
|
2nd season |
129 |
183 |
233 |
280 |
210 |
|
Total |
307 |
404 |
664 |
830 |
550 |
|
Length of growing season (months) |
|
|
|
|
|
|
1st season 0.5 |
1.4 |
2.4 |
|
2.6 |
1.9 |
|
2nd season 0.5 |
0.8 |
1.2 |
|
2.2 |
1.1 |
|
Total 1.0 |
2.2 |
3.6 |
|
4.8 |
3.0 |
|
Forage yield (t DM/ha) |
|
|
|
|
|
|
1st season |
0.2 |
1.0 |
2.0 |
2.2 |
1.5 |
|
2nd season |
0.2 |
0.4 |
0.8 |
1.7 |
0.8 |
|
Total |
0.4 |
1.4 |
2.8 |
3.9 |
2.3 |
|
No. of years |
3 |
12 |
10 |
5 |
30 |
1 Very low = < 1 t DM/ha per year; low = 1.0-2.0 t; medium = 2.1 -3.4 t; high = > 3.4 t.
Table 10.5. Characterisation of cattle productivity parameters for year-types grouped by resource class.
|
|
Resource class1 |
Mean |
||||
|
Very low |
Low |
Medium |
High |
|||
|
Calving (%) |
36 |
54 |
54 |
48 |
51 |
|
|
Milk yield per cow with calf |
113 |
190 |
268 |
348 |
234 |
|
|
Liveweight |
169 |
183 |
196 |
211 |
190 |
|
|
Mortality (%) |
|
|
|
|
|
|
|
|
Cows |
40.0 |
9.1 |
5.2 |
5.4 |
10.3 |
|
|
Stock <2 years |
45.0 |
11.9 |
8.1 |
7.8 |
12.2 |
|
|
Stock 2-3 years |
25.0 |
7.6 |
5.2 |
5.2 |
8.2 |
|
|
Stock >3 years |
18.3 |
6.4 |
4.9 |
4.6 |
6.8 |
1 Very low = < 1 t DM/ha per year; low = 1.0-2.0 t; medium = 2.1 -3.4 t; high = > 3.4 t.
The Maasai occasionally bring into their herds heifers, bulls and steers they obtain by exchange or purchase and a provision was made in the model for such acquisitions. Again, the number acquired was fixed as observed during the study period, except that none were acquired during drought periods.
10.2.1 Herd size and stocking rate
10.2.2 Herd productivity
10.2.3 Milk offtake
10.2.4 Net output
10.2.5 Effects of increased offtake of steers on herd and ranch productivity
The modelled long-term fluctuations of population in the three herd sizes and for the entire Olkarkar ranch are shown in Figure 10.6. Two cycles of herd growth and decline are apparent.
In general, the mean rate of herd decline during drought periods was 14% per year. Thus if a drought persists for 2 years the cattle population will be reduced by 26%. If the drought continues for a third year the herd size will decline to 63% of its pre-drought level. In the serious drought that occurred in years 27 and 28 the cattle population was reduced to 68% of its pre-drought level in only 2 years. Mean herd growth during the recovery periods was 7.5% per annum.
Forage supplies fluctuate more rapidly and more widely than the cattle population, hence imbalances between available grazing resources and cattle population can be expected. The magnitude and duration of periods of overstocking and understocking depend on the average herd size and the assumed safe stocking rate.
A safe stocking rate was calculated by assuming a daily forage demand of 10 kg DM/TLU or a rate of utilisation of about 60% of the standing herbage biomass, given a daily intake of 2.5% of bodyweight or 6.25 kg DM (see Section 4.4.3: Carrying capacity). Individual years do not occur in isolation as there is a carry-over of forage supplies from the previous to the current year. Thus, moving averages over 2 years were used to estimate the safe stocking rate. The livestock biomass in TLU for the entire ranch in each year was derived from the mid-year aggregated herd size, its age/sex/class composition and the liveweight of each class.
The long-term balance between forage supply and stocking rate for the 10000-ha ranch is shown in Figure 10.7. This shows a pattern of periods of understocking alternating with periods of overstocking. During drought periods, the amount of forage available fell to 4.5-5.7 kg DM/TLU per day, which is less than the minimum required intake. However, the ranch was correctly stocked or understocked for 22 out of 30 years, and was seriously overstocked for only 5 years. Over the entire 30-year period, forage supply and demand were in balance, with both the safe stocking rate and the herd size showing a median value of 5600 TLU for the 10000-ha ranch. Given the fairly conservative forage utilisation rate adopted, it can be concluded that the long-term carrying capacity of the ranch was about 0.6 TLU/ha (1.7 ha/TLU), which is similar to the actual stocking rate of Olkarkar ranch during the 1981-83 period (see Section 5.3.2: Grazing patterns end stocking rates in the northern ranches).
Herd productivity can be measured in several ways, including stock biomass production, milk offtake net output expressed in monetary terms and rates of return on labour, land and capital invested in livestock. These measures are largely influenced by herd size, which fluctuates from year to year. The overall productivity of the ranch was dominated by the dynamics of the large herds belonging to the rich producers as these constitute nearly 80% of the total cattle population of the ranch. Although, proportionally, changes in the herd sizes of the poor and medium-wealth producers were more pronounced than changes in large herds, their effect on the fluctuations in the total ranch came population was minimal. In the three droughts that occurred during the 30 years modelled, poor producers lost an average of 43% of their herds during each drought, medium-wealth producers lost 39%, while rich producers lost only 34%. The poor producers had proportionally more cows and calves in their herds than did medium-wealth and rich producers, and these classes of stock were more likely to die during drought than other stock classes (Table 10.3).
Biomass production
Cattle biomass production is defined as the total change in herd biomass during the year. It includes the weight gain of all classes of animals remaining in the herd at the end of the year plus the weight of animals sold and slaughtered for home consumption. In normal years this is a positive value, but was negative in drought years because of high mortality rates and weight losses.
The simulated long-term (30-year) mean annual liveweight production for both the poor and medium-wealth producers was 43 kg/TLU, compared with only 19 kg/TLU for rich producers (Table 10.6). This is explained by the low level of offtake particularly sales, practised by rich producers (Table 10.7). The low sales offtake of the rich producers depressed liveweight production per TLU for two reasons: first, animals did not gain much weight beyond the age of 5 years and low sales resulted in an increase in the proportion of older animals in rich producers' herds; and second, many of the animals accumulated in good years died or lost weight during drought periods.
Simulated mean liveweight production for Olkarkar as a whole was 24 kg/TLU (13 kg/ha), ranging from a loss of 102 kg/TLU (-30 kg/ha) in drought years to a gain of 42 kg/TLU (42 kg/ha) in the best years (Table 10.6).
The mean annual liveweight production of 13 kg/ha compares favourably with the 9 kg produced by Boran pastoralists in southern Ethiopia and the 4.3 kg produced on Australian cattle stations (Cossins and Upton, 1987), but is considerably less than that achieved on some commercial ranches in Kenya.
Table 10.6. Simulated long-term livestock productivity of poor, medium-wealth and rich producers and for the ranch as a whole under different year-types.
|
Period type |
No. of years |
Livestock productivity |
||||
|
kg/TLU per year |
Ranch |
kg/ha per year |
||||
|
Wealth class1 |
||||||
|
Poor |
Medium |
Rich |
||||
|
Long term |
30 |
43 |
43 |
19 |
24 |
13 |
|
Drought |
3 |
- 127 |
-96 |
-101 |
-102 |
-30 |
|
Poor |
7 |
35 |
31 |
10 |
15 |
75 |
|
Fair |
9 |
55 |
50 |
28 |
33 |
17 |
|
Good |
8 |
57 |
55 |
31 |
37 |
21 |
|
Best |
3 |
61 |
60 |
37 |
42 |
42 |
|
Study period |
(1981-83) |
73 |
74 |
48 |
54 |
33 |
1 Poor = <5 tropical livestock units (TLU) per active adult male equivalent (AAME); medium = 5-12.99 TLU/AAME; rich = ³ 13 TLU/AAME.
Table 10.7. Annual sales offtake by poor medium-wealth and rich producers under different year-types.
|
Wealth class1 |
Drought years |
Best years |
Long term |
|||
|
Offtake in per cent of |
Offtake in per cent of |
Offtake in per cent of |
||||
|
No. |
Biomass |
No. |
Biomass |
No. |
Biomass |
|
|
Poor |
15 |
19 |
11 |
13 |
12 |
16 |
|
Medium |
15 |
19 |
11 |
14 |
12 |
17 |
|
Rich |
6 |
8 |
5 |
6 |
5 |
7 |
1 Poor = <5 tropical livestock units (TLU) per active adult male equivalent (AAME); medium = 5-12.99 TLU/AAME; rich = ³ 13 TLU/AAME.
The modelled results of milk availability for human consumption showed wide fluctuations across years. The long-term mean availability of milk for poor and medium-wealth producers was 1563±143 and 2348±211 kg per household per year respectively (Table 10.8). In most years poor producers did not produce enough milk to meet their target of obtaining 65-70% of their energy from milk (Nester, 1985). Rich producers had far more milk than their households needed in all years except during the first drought, when they had only 1992 litres of milk available, compared with the long-term average of 10836±968 litres per year. Since there was no ready market for the excess milk of rich producers it was largely left for the calves. Rich producers also gave milk to poorer relatives and friends. For purposes of economic analysis, only the production of 5000 litres of milk per year is assumed to have economic value.
Table 10.8. Simulated milk offtake medium-wealth and rich producers and for the ranch as a whole under different year-types.
|
Period type |
No. of years |
Annual milk offtake (litres/household) |
Group ranch |
|||
|
Wealth class1 |
||||||
|
Poor |
Medium |
Rich |
Litres/TLU |
Litres/ha |
||
|
Long term |
|
1563 |
2348 |
5000a |
24 |
12 |
|
Drought |
2 |
565 |
825 |
3525 |
25 |
7 |
|
Poor |
7 |
1090 |
1663 |
5000 |
22 |
11 |
|
Fair |
9 |
1488 |
2262 |
5000 |
23 |
12 |
|
Good |
8 |
2116 |
3143 |
5000 |
24 |
14 |
|
Best |
3 |
2415 |
3608 |
5000 |
22 |
15 |
|
Study period |
(1981-83) |
2480 |
3550 |
5000 |
26 |
15 |
1 Poor = < 5 tropical livestock units (TLU) per active adult male equivalent (AAME); medium = 5-12.99 TLU/AAME; rich = ³ 13 TLU/AAME.a Only the first 5000 litres of production was considered.
Milk availability per person in households of different wealth class is shown in Table 10.9. Rich producers have more than enough milk for their household (target of about 360 litres/active adult male equivalent (AAME); in all years except during droughts, when milk availability dropped below 200 litres/AAME. In contrast, medium-wealth producers achieved the target level of production only in good and the best years and poor households only in the best years.
The net values of output for the three types of producers were computed using constant 198183 prices (Table 10.10). The long- term mean annual net output per household of large-scale producers was 3.3 times that of poor producers and 2.3 times that of the medium-wealth producers. However, these differences narrowed to 2.0 and 1.9 times respectively when expressed on a per caput basis because of the larger number of people in rich households.
During drought years all producers sustained a net loss of output, with rich households suffering much greater losses than poor and medium wealth households (5.6 and 3.9 times as large, respectively, on a per caput basis). In contrast, the net output of rich producers in the best years was only 2.4 times that of poor producers and 2.1 times that of medium-wealth producers on a per caput basis.
Table 10.9. Simulated milk offtake per person by poor, medium-wealth and rich households under different year-types.
|
Period type |
Milk offtake (litres/AAME1) |
||
|
Wealth class2 |
|||
|
Poor |
Medium |
Rich |
|
|
Long term |
233 |
272 |
> 500 |
|
Drought |
84 |
96 |
191 |
|
Poor |
162 |
193 |
> 500 |
|
Fair |
221 |
262 |
> 500 |
|
Good |
315 |
365 |
>500 |
|
Best |
359 |
419 |
> 500 |
|
Study period |
370 |
507 |
> 500 |
1 Active adult male equivalent.2 Poor = <5 tropical livestock units (TLU) per active adult male equivalent (AAME); medium = 5-12.99 TLU/AAME; rich = ³ 13 TLU/AAME.
The long-term mean net output for Olkarkar as a whole was KSh 59/ha per year or KSh 1535 per person. The net loss during severe drought periods was KSh 109/ha and KSh 2645 per person. During the best years net output per person was 2.4 times the long-term mean.
A comparison of net returns accruing to capital invested in livestock for the three producer wealth classes and for the ranch as a whole during three year-types is shown in Table 10.11. Again, productivity was inversely related to wealth class. The long- term mean net return was 17%, ranging from 9% for rich producers to 24% for poor producers.
Table 10.10. Simulated net output of poor, medium-wealth and rich producers and the ranch as a whole under different year-types.
|
Wealth class1 |
Net output (KSh/year) |
|||
|
Year-type |
Study period |
|||
|
Long term |
Drought |
Best |
||
|
Per household |
|
|
|
|
|
Poor |
7425 |
6397 |
13827 |
12990 |
|
Medium |
10309 |
-11800 |
19761 |
24075 |
|
Rich |
24495 |
-58708 |
53513 |
60880 |
|
Weighted mean |
17463 |
-24925 |
33725 |
33260 |
|
Per person |
|
|
|
|
|
Poor |
1105 |
-952 |
2058 |
1930 |
|
Medium |
1196 |
-1369 |
2292 |
2790 |
|
Rich |
2237 |
-5362 |
4887 |
5560 |
|
Weighted mean |
1535 |
-2645 |
3753 |
3790 |
|
Per TLU |
|
|
|
|
|
Poor |
238 |
437 |
322 |
380 |
|
Medium |
184 |
-345 |
268 |
320 |
|
Rich |
86 |
-342 |
149 |
195 |
|
Weighted mean |
168 |
-377 |
245 |
152 |
|
Per ha |
|
|
|
|
|
Ranch |
59 |
-109 |
122 |
230 |
1 Poor = < 5 tropical livestock units (TLU) per active adult male equivalent (AAME); medium = 5-12.99 TLU/AAME; rich = ³ 13 TLU/AAME;
The high net returns realised by poor and medium-wealth producers were the result of their intensive milking practices. As was noted earlier, rich producers extracted less than 40% of the milk potentially available and their long-term annual offtake of animals was only 5%. The productivity of rich producers could be markedly increased by increasing their offtake of both milk and animals. However, there was no ready market for milk in the study area. The effects of higher offtake rates of animals for sale by medium-wealth and rich producers is discussed in the next section.
Table 10.11. Simulated net return on capital invested in livestock of poor, medium-wealth and rich producers as a who/e under different year-types.
|
Wealth class1 |
Net return on capital invested in livestock (%) |
|||
|
Year-type |
Study period |
|||
|
Long term |
Drought |
Best |
||
|
Poor |
24 |
-32 |
34 |
39 |
|
Medium |
18 |
-30 |
28 |
32 |
|
Rich |
9 |
-30 |
16 |
21 |
|
Weighted mean |
17 |
-31 |
26 |
25 |
1 Poor = < 5 tropical livestock units (TLU) per active adult male equivalent (AAME); medium = 5-12.99 TLU/AAME; rich = ³ 13 TLU/AAME;
Pastoralists tend to keep their herds as large as practically possible as a way of coping with the effects of droughts, on the basis that the larger one's herd at the beginning of a drought, the more likely one will have a viable herd at the end of the drought. However, pastoralists often delay selling stock as long as possible, with the result that the animals when sold are in very poor condition and fetch very low prices. Furthermore, flooding of the market with such animals also severely taxes the capacity of the market to absorb the increased supply. Consequently, many animals die despite pastoralists' belated willingness to sell in distress (Grandin and Lembuya, 1987). This results in a considerable economic loss both to the producers and the nation. One way of avoiding such losses is to increase sales of animals during favourable periods.
Since steers are not part of the breeding herd, their presence or absence does not affect the regeneration of the herd after drought or milk supplies. It was therefore postulated that increased offtake of steers would not reduce herd viability. The long-term productivity analysis kept sales of animals constant at 4, 7 and 17 head for poor, medium-wealth and rich producers respectively. A sensitivity analysis was performed using the long-term herd-projection model to determine the effect of a higher level of steer offtake on herd productivity. In the high-level offtake model, all steers of the medium-wealth and rich producers were sold upon reaching 5 years of age, in addition to the cull cows and bulls ordinarily sold.
The results indicate that there was lime scope for the medium-wealth producers to increase their sales offtake from the 7 head per year they sold during the study period. There were only 2 years out of the 30 that sales of steers could be increased, and then only to 8 head in one year and 9 head in the other.
In contrast, rich producers could increase their sales in 25 of the 30 years modelled and could achieve a mean sales offtake of 25 head per year. This represents a 47% increase in the sales offtake of this class of producer.
The aggregate result of such a policy of increased sales offtake of steers would be to increase the long-term mean sales of the ranch from 395 to 510 head per year. Table 10.12 shows that such a sales policy could substantially increase the long- term annual productivity of both the rich producers and the whole group ranch. It would also reduce grazing pressure on the ranch by reducing the mean cattle population by 19% to 4692 head, which is about the 1981-83 level of stocking on Olkarkar. Increased offtake increased liveweight production on the ranch by about 80% per TLU and 30% per ha (Table 10.12). The return on capita' invested in livestock increased from 9% to 14% per annum for the rich producers and from 11 % to 16% per annum for the ranch as a whole (Table 10.13). The discounted net output over the whole 30-year period was increased by 29% for the rich producers and by about 19% for the whole ranch.
Conclusion
On the whole, poor producers with 30 head of cattle extracted as much milk and meat as possible from their cattle Their long- term animal offtake was about 16% of biomass, compared with only 7% for rich producers with 300 cattle or more. In terms of milk offtake across the entire period the rich producers extracted about 70% of the potential of their cows, compared with nearly 100% by the poor producers. The aggregate result of the high exploitation of production by the poor producers was a long-term mean return on their capital in livestock of 24% p.a., compared with a mere 9% for rich producers.
Table 10.12. Impact of increased sales offtake on annual herd productivity of rich producers on Olkarkar Group Ranch and of the ranch as a whole.
|
Parameter |
Sales offtake |
Ranch |
|||||
|
Rich producers1 |
|||||||
|
Normal |
Increased |
Change (%) |
Normal |
Increased |
Change |
||
|
No. of animals sold |
17 |
25 |
70 |
395 |
510 |
28 |
|
|
No. of animals died |
36 |
32 |
-9 |
691 |
626 |
-10 |
|
|
Herd size (head) |
392 |
312 |
-20 |
5776 |
4692 |
-19 |
|
|
Stocking rate (ha/TLU) |
|
|
|
1.9 |
2.3 |
21 |
|
|
Liveweight offtake |
|
|
|
|
|
|
|
|
|
kg/TLU |
39 |
72 |
85 |
20 |
36 |
80 |
|
|
kg/ha |
|
|
|
11 |
14 |
30 |
1 Rich = ³ 13 tropical livestock units (TLU) per active adult male equivalent (AAME);
Table 10.13. Impact of increased sales offtake long-term annual net output of rich producers on Olkarkar Group Ranch and of the ranch as a who/e.
|
Parameter |
Sales offtake |
|||||
|
Rich producers1 |
Ranch |
|||||
|
Normal |
Increased |
Change (%) |
Normal |
Increased |
Change (%) |
|
|
Net output |
|
|
|
|
|
|
|
KSh/household |
24495 |
29775 |
12 |
17463 |
18640 |
9 |
|
KSh/caput |
2237 |
2719 |
12 |
1635 |
1823 |
9 |
|
KSh/TLU |
86 |
138 |
62 |
168 |
184 |
45 |
|
KSh/ha |
|
|
|
59 |
66 |
29 |
|
Return to capital invested in livestock (%) |
9 |
14 |
56 |
11 |
16 |
50 |
|
Discounted net output @ 12% p.a. (KSh '000) |
189 |
243 |
29 |
4208 |
5021 |
19 |
1 Rich = ³ 13 tropical livestock units (TLU) per active adult male equivalent (AAME).
The low rate of return obtained by owners of large herds is explained by the fact that up to 55% of their annual biomass production is saved in the form of stock accumulation, much of which is lost when major droughts occur. This implies that the scope for increasing the productivity of rich households, which constitute 40% of the human population of the ranch but control nearly 80% of the livestock biomass, does not lie in improved technology but rather in greater exploitation of what is already being produced. On the other hand, the livestock productivity of poor households could be increased only by intensifying production via forage conservation, establishment of feed gardens, improved calf rearing and animal health care (see Section 11.2: The improvement of cattle productivity).
Cossins J J and Upton M. 1987. The Borana pastoral system of southern Ethiopia. Agricultural Systems 25:199-218.
Grandin B E and Lembuya P 1987. The impact of the 1984 drought in Olkarkar Group Ranch, Kajiado, Kenya. ODI Pastoral Development Network Paper 23e. Overseas Development Institute, London, UK. 14 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.
de Leeuw P N. 1986. Modelling pastoral livestock production: Problems and prospects. In: de Ridder N. van Keulen H. Seligman N G and Neate P J H (eds), Modelling of extensive livestock production systems. Proceedings of the ILCA/ARO/CABO workshop held at Bet Dagan, Israel, 5-9 February 1985. ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia. pp. 250-273.
de Leeuw P N and Konandreas P A. 1982. The use of an interger and stochastic model to estimate the productivity of four pastoral production systems in West Africa. In: Proceedings of the National Conference on Beef Production. July 1982, Kaduna, Nigeria. pp. 184-212.
de Leeuw P N and Nyambaka R. 1988. The prediction of rangeland production from rainfall data in arid and semi-arid eastern Africa. In: Pasture Network for Eastern and Southern Africa (PANESA), African forage plant genetic resources, evaluation of forage germplasm and extensive livestock production systems. Proceedings of the Third Workshop held at the International Conference Centre, Arusha, Tanzania, 27-30 April 1987. ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia. pp. 260-268.
Musembi D K. 1984. Identification of crop-growing seasons of semi-arid Kenya by analysis of the soil moisture patterns. MSc thesis, Texas A & M University, Texas, USA. 129 pp.
Musembi D K. 1986. The seasonal climate of rangelands. In: Hansen R M, Woie B M and Child R D (eds), Range development and research in Kenya. Proceedings of a conference held at the Agricultural Resources Centre, Egerton College, Njoro, Kenya, 1 -5 April 1986. Winrock International Institute for Agricultural Development, Morrilton, Arkansas, USA. pp. 183-200.
Nestel P S. 1985. Nutrition of Maasai women and children in relation to subsistence food production. PhD thesis, Queen Elizabeth College, University of London, London, UK. 223 pp.
Potter H L. 1985. Aspects of climate, herbage growth and animal production in a semi-arid area of Kenya. PhD thesis, University of Nairobi, Nairobi, Kenya. 195 pp.
Sullivan G M, Cartwright T C and Farris D E. 1981. Simulation of production systems in East Africa by use of interfaced forage and cattle models. Agricultural Systems 7:245-265.
Wagenaar K T and Kontrohr E. 1986. Appraisal of the ILCA cattle herd dynamics model using data from pastoral systems in Mali and Kenya. In: de Ridder N. van Keulen H. Seligman N G and Neate P J H (eds), Modelling of extensive livestock production systems. Proceedings of the ILCA/ARO/CABO workshop held at Bet Dagan, Israel, 5- 9 February 1985. ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia. pp. 231-249.
