Livestock Research for Rural Development 21 (9) 2009 Guide for preparation of papers LRRD News

Citation of this paper

Modelling of long term pasture production and estimation of carrying capacity of Ankole pastoral production system in South Western Uganda

H Mulindwa1, 3, E Galukande2, 3 M Wurzinger3, 4, A Okeyo Mwai4 and J Sölkner3

1National Livestock Resources Research Institute, Tororo, Uganda
mulindwaha@yahoo.com
2National Animal Genetic Resources Center and Data Bank, Entebbe, Uganda
3BOKU - University of Natural Resources and Applied Life Sciences, Vienna
4International Livestock Research Institute, Nairobi, Kenya

Abstract

The Ankole pastoral production system in South Western Uganda is based on grazing without supplementary feeding.  A stochastic simulation model was developed to determine the dynamics of pastures grazed by Ankole cattle and their Holstein Friesian crosses and the carrying capacity (CC) of the livestock grazing system. The model used the concept of rain use efficiency which relates pasture production to rainfall. A cross sectional study was carried out on 16 selected farms and data on number of animals, sex, age group and size of available grazing land was collected.

 

The similarity between the results of the simulation rainfall runs and field data are considered to be satisfactory. The overall annual forage production is 3905 ± 72kg/ha. The lowest CC (5.65 ± 0.75) occurs in long dry season (June to August) while the highest CC (1.41 ± 0.06 ha/TLU) occurs in short rain season (September to November). Annual carrying capacity ranges between 1.88 and 2.08 ha/TLU with an overall mean of 1.95 ± 0.04 ha/TLU. Sixty three (63%)  percent of the surveyed farms have stocking rates that are higher than the CC throughout the year while the rest are overstocked in the dry seasons of the year.

 

The results indicate that CC is dynamic and its variability is more pronounced within the year than between years. In response to seasonal CC, the major point of intervention in regard to reduction of actual stocking rates could be done in May shortly before the start of the long dry season. For Ankole pastoral system to be sustainable, the stocking rate should not go below 1.41 ha/TLU.

Key words: Rain use efficiency, rangelands, simulation, stocking rate


Introduction

The Ankole cattle production system in South Western Uganda is based on grazing without supplementary feeding. In this area livestock production is highly dependent on the availability of pastures, the quantity and quality of which are primarily determined by the amount and distribution of rainfall. This rainfall pattern formerly forced cattle keepers to move their cattle from place to place in search of water and pastures. However, the system is undergoing transformation from nomadic to sedentary system due to a new national policy that allows change of land tenure from communal to private rangeland tenure (Sserunkuuma and Olso 1998, and Kisamba et al 2006). Another on-going trend is the practice whereby pastoralists raise two separate herds on the same land, an improved Ankole-Friesian herd and pure Ankole as a risk management strategy, in order to enhance the productivity of their herds. This practice results in high stocking densities and higher likelihood of overgrazing. In a study by Sserunkuuma and Olso (1998) in South Western Uganda, the authors concluded that any program aimed at enhancing range productivity and mitigating overgrazing should attach as much weight to feeding and animal health care, as it does to breed improvement. A wide variation in pasture quality and quantity, as well as cow performance was observed between seasons (Okello et al 2005). The success of the current grazing strategy will therefore depend on the ability to track forage availability on the range and matching it to the number of animals that can be grazed on the rangeland. The amount of available forage and the number of animals grazing on an area affect intake and therefore animal nutritional performance, productivity per unit area and the long term ecological health of the rangeland.

 

The close relation between annual rainfall and rangeland productivity is widely acknowledged e.g. Coe et al 1976, De Leeuw and Nyambaka 1987. As water is the main limiting factor in most rangelands, high and well distributed rainfall will result in increased productivity. Climate variables, especially rainfall in semiarid and arid areas, have overriding effects on grassland production, and thus affect livestock carrying capacity (Mei et al 2004). Carrying capacity refers to the total number of animals that may be safely supported by a unit rangeland in the long term (Caltabiano 2006). A predictive model for carrying capacity, determined directly from rainfall, was established by Coe et al (1976) based on data obtained mainly from natural ecosystems in Africa. Oesterheld et al (1992) developed separate regression models between aboveground net primary production and herbivore biomass for wildlife and livestock based on data from South America. Carrying capacity, or the ability of a given area to support a certain population of animals on a continuing basis (De Vos 1969) may be altered by both long and short term variations in climate and particularly in precipitation (Phillipson 1975). The objective of this study was to develop a computer simulation model to predict the dynamics of herbage productivity of the range grazed by Ankole cattle and crosses with Holstein Friesian and to determine the carrying capacity of the rangeland.

 

Materials and methods 

The study was carried out in Mbarara district of South Western Uganda. The topography consists of undulating hills and valleys. The hills rise about 100 – 200 m above the flat valley bottoms. The area lies between 1250 – 1525 m above sea level. Rainfall has in the past been reliable, but recent trends in rainfall patterns indicate more erratic behavior. The rainfall occurs in a bimodal pattern, peaking in the months of April to May and September to November. The current study used historical rainfall data for 46 years and obtained an annual rainfall mean of 939 mm. The months of June, July and August normally constitute a dry season with no rainfall. Schwartz et al (1996) have computed the mean annual rainfall for the Mbarara Meteorological site at Kakoba for the period 1980 -1994 and obtained a figure of 882 mm with a coefficient of variation of 20%. Mean maximum temperature is about 26º C and mean minimum around 14º C. Themeda triandra, Cynodon dactylon, Panicum maximum, Brachiaria decumbens,B. platynota and Chloris gayana are quantitatively the most valuable forage grasses for cattle. Hyparrhenia filipendula, Loudentia kagerensis, Digitaria maitlandi and to a small extent, Sporobolus pyramidalis, are less important due to their lower quality, but form a major component of the grass vegetation (Okello et al 2005).

 

Model description

 

A dynamic stochastic compartment model based on difference equations programmed in STELLA 9.0.2, 2007 (High Performance Systems, Inc., Hanover, New Hampshire,) was developed. The simulations are based on a one-month time step (units of the model are in months). Due to the stochastic nature of the model, results of the simulation are presented as mean for 50 separate runs of the model. The main assumption in the model is that vegetation growth is directly related to the rainfall dynamics. The model simulates the dynamics of standing green forage using the concept of rain use efficiency (RUE, kg DM /ha /mm/year). The RUE factor, which refers to a relationship between maximum standing crop at the end of a rainy season, and total annual rainfall (mm), was used to calculate the carrying capacity of Ankole pastoral system. Le Houérou et al (1988) reported a RUE = 4·0 ± 0·3 for range type, condition and productivity for areas with similar conditions to those of Ankole pastoral production system. In order to simulate the observed seasonal variation in forage production, monthly rainfall is generated randomly from a cumulative relative frequency distribution (Grant et al 1997) for each month, created from real system historical rain fall data of 46 years (from 1961 to 2007) obtained from the Kakoba Metrological Department, Mbarara.

 

Equation 1  

 

MRF = RANDOM (Cumulative relative frequency)

Where, MRF is the monthly rainfall (mm) and cumulative relative frequency is a value picked randomly from 0-1.

 

Random is a built-in function in STELLA that generates a series of uniformly distributed random numbers between 0 and 1 for each month and samples a new random number in each iteration of a model run. Each randomly picked number has got a corresponding amount of rainfall depending on the month (January to December) being simulated. Pasture growth is simulated in monthly time step using a multiplicative function of rain use efficiency and monthly rainfall. Seasonal forage growth was obtained by summation of forage growth that occurred during the months that constitute a particular season.  The four seasons considered were short dry season (December-February), long rain season (March to May), long dry season (June to August) and the short rain season (September to November). The model input data required, consisted of climate and intake of tropical livestock unit (TLU). Annual rainfall is modeled as the summation of the individual monthly rainfall within a given year.

 

Equation 2

 

GR = RUE * MRF,

where GR is the monthly forage growth (kg DM/month/ ha)

 

A basic technique for determining carrying capacity is to calculate the total amount of forage at the end of the growing season, multiply this by a correction factor and then divide by the average yearly feed requirements of a livestock unit (Hocking and Mattick 1993). Not all range forage can be used by livestock, some is not accessible to the animals and some is unpalatable and further losses occur due to senescence and by trampling by the animals (Hocking and Mattick 1993). In order to account for sustainability a proper use factor was included, which varies according to different researchers and different situations from 30% in Southern Ethiopia (Cossins and Upton 1987) to 45% in Tsavo, Eastern Kenya (Van Wijngaarden 1985). Van Wijngaarden (1985) estimates that not more than 55% of the grass cover should be removed in one way or the other to keep the grasslands at least in the same condition as it was before. So, if utilizing the grasslands should be sustainable, to prevent degradation, at least 45% of the peak standing crop should be left at the beginning of the next rainy season. Other authors, Kavana et al (2005) and Mugerwa (1992) used a proper use factor of 50% while Caltabiano (2006) proposed a factor of 30% for black spear grass and 20 % for mulga pastures. In this study, a year long proper use factor of 30% as used by Guevara et al (1996) was adopted and used as consumable forage. The model uses the measure of livestock input on the range known as the tropical livestock unit (TLU) to calculate the carrying capacity (CC) of the range. Tropical livestock unit is a unit representing a ruminant of 250 kg live weight (Sserunkuuma and Olson 1998). Daily feed intake per TLU was taken at 2.5% of body weight.  The following equation for the use of forage production was applied (FAO 1991):

Carrying capacity (CC) = animal requirement / weight of standing crop * proper use factor

 

A cross sectional study was carried out in April 2006 on 16 selected farms. The observations recorded included number of animals, sex, age group and size of grazing area. All cattle age groups were converted to Tropical Livestock Units (TLU). A Tropical Livestock Unit is a hypothetical animal of 250 kg live weight. TLU is used to bring all animal types under a common denominator (using conversion factors: 0.25 TLU for calf, 0.5 TLU for Ankole heifer, 0.6 for Ankole-Friesian heifer, 1 TLU for  Ankole cow and 1.2 TLU for Ankole-Friesian cow and 1.5  TLU for bulls respectively).

 

Results and discussion 

The similarity between the results of the simulation rainfall runs and field data are considered to be satisfactory (Figure 1). 



Figure 1.  Actual and simulated monthly rainfall


The simulated annual mean rainfall was 976 mm which is comparable to the actual annual mean of 939 mm. In Table 1, the mean long-term annual forage productivity was predicted to be 3905 ± 73 kg/ha over a 30 year period which is close to 3900 kg/ha reported by Mugerwa (1992) and considerably lower than 4560 kg/ha estimated by Byenkya (2004).


Table 1.  Pasture production, rainfall and carrying capacity in Ankole pastoral system

Year

Mean forage production,

kg DM/ha/year

Mean annual rainfall,
mm

Carrying capacity, 
ha/TLU

1

3811

953

2,00

2

3664

916

2,08

3

3970

992

1,92

4

3865

966

1,97

5

3878

969

1,96

6

3943

986

1,93

7

3977

994

1,91

8

3829

957

1,99

9

3850

962

1,98

10

3974

994

1,91

11

3897

974

1,95

12

3858

965

1,97

13

4036

1009

1,88

14

3921

980

1,94

15

3932

983

1,93

16

3951

988

1,92

17

3902

976

1,95

18

3900

975

1,95

19

3936

984

1,93

20

3876

969

1,96

21

4014

1004

1,89

22

3927

982

1,94

23

3876

969

1,96

24

3849

962

1,98

25

3920

980

1,94

26

3826

957

1,99

27

3886

972

1,96

28

3988

997

1,91

29

3921

980

1,94

30

3977

994

1,91


In this study, the minimum and maximum forage produced were 3664 kg/ha and 4036 kg/ha respectively.  The difference between the current study and previous ones could be due to the fact that earlier studies were based on one-year rainfall whereas the current study is based on 30 year rainfall data. Predictions of dry matter have been overestimations and are often unreliable when made for one particular year (FAO 1991). The overall carrying capacity predicted in this study was 1.95 ± 0.04 ha/TLU which is higher than 2.27 ha/TLU (Byenkya 2004) and lower than 1.63 ha/TLU reported by Mugerwa (1992). Hocking and Mattick (1993) reported a carrying capacity in the range 2.5 - 3.5 ha/LU for wooded grasslands of Tanzania receiving 875 -1000 mm of annual rainfall. Carrying capacity variability is more pronounced within year than between years (Tables 1 and Figure 2).



SDS = Short dry season (December to Febuary), LWS = Long wet season (March to May),

LDS = Long dry season (June to August), SWT = Short wet season (September to November)


Figure 2.  Mean seasonal carrying capacity


The lowest CC (5.65 ± 0.75) occurs in long dry season (June to August) while the highest CC (1.41 ± 0.06 ha/TLU) occurs in short rain season (September to November). The dynamic nature and seasonal changes are dramatically visible though the variability of annual carrying capacity was negligible ranging between 1.88 ha/TLU and 2.08 ha/TLU.  The results show that carrying capacity is a dynamic concept requiring active monitoring and rapid adjustments of stocking rates. The implication of this could be a move towards more flexible and short-term responses to environmental variation. Currently farmers respond to within year changing carrying capacity, especially in the dry season, by transferring pure Ankole herds to distant alternative rangelands or grazing in adjacent unfenced land. However, this practice will soon not be possible because of increased demand for land due to population increase.

 

The observed stocking rates are presented in Table 2.


Table 2.  Tropical livestock units, grazing area and observed stocking rate on individual farms

Farm

TLU

Grazing area, ha

Observed Stocking density, ha /TLU

1

291

215

0.74

2

135

120

0.89

3

252

366

1.45

4

102

108

1.05

5

559

200

0.36

6

151

100

0.66

7

131

221

1.68

8

72

150

2.07

9

281

360

1.28

10

301

300

1.00

11

296

270

0.91

12

152

122

0.80

13

369

750

2.03

14

124

240

1.93

15

218

350

1.60

16

185

220

1.27


Basing on results in figure 2 and Table 2, it is clear that 63% of the surveyed farms have stocking rates that are higher than the CC throughout the year while 37% overstock in the dry months of the year which indicates a risk of overgrazing in this production system. Traditionally, in the study area cattle are a store of wealth or savings from which withdraws are made only for special social occasions or emergency needs hence a reluctance for pastoralists to sell cattle regularly even when faced with forage scarcity.

 

Although famers keep stoking rates that are higher than the carrying capacity, it does not necessary mean that animals do not meet their nutrient requirements. The CC obtained in this study was based on dry matter yield without considering  nutrient quality of pasture and yet Okello et al (2005) reported crude protein content of cattle diets that showed a third peak in August (60g kgDM–1) in addition to the May and November peaks (100 and 80g kgDM–1, respectively).  Dietary crude protein peaks at the climax of herbage growth and falls in the late wet season (Okello et al 2005) and through dietary selection as well as increased browsing activity of cattle in the dry season animals are able to meet the 60–80g kgDM–1 crude protein level required for optimal digestion and feed intake (Minson 1981) and hence compensate for low forage crude protein in the dry season and low available forage per TLU caused by high stocking rates. Nevertheless, according to Okello et al 2005, live weight increased with each rainy season, peaking in May and November, before consistently declining to the lowest levels in the dry seasons. Farmers need to make marginal adjustments to actual stocking rates in response to seasonal carrying capacity and the major point of intervention could be done in May, shortly before the start of long dry season. However, massive reduction in stocking rates, at a specific time of the year, through off-takes could lead to low prices for the sold animals as a result of demand and supply forces. In the short term, stocking rate reductions could be done in two ways namely by off-takes and transferring animals to alternative rangelands. As a long-term strategy, improvement and establishment of marketing infrastructure and institutions such as regular cattle auction houses, slaughter houses and processing meat plants could play a role in keeping the prices stable and encourage managed off-takes and restocking of animals in response to biological variation of the range carrying capacity.

 

Government intervention aimed at changing the attitude of farmers through sensitization on the dangers of overstocking could help farmers adopt sustainable stocking rates. The sensitization could be easily integrated in the current on-going national program; national agricultural advisory services (NAADS). The current study has determined the extreme potential carrying capacity values (1.41 and 5.65 ha/TLU), there is need to evaluate selected stocking rates that fall within the above range for economic viability and the findings of such a study will provide  economically profitable range of stocking rates within a cattle keeper could operate other than recommending a single static CC because a range ensures social interests (keeping larger herds) as well as economic viability with ecological sustainability. Operating within a range of stocking rates will occasionally lead to slight overgrazing in some months causing loss of body weight but the effect will not be severe because cattle will be able to recover in the more favourable months. The practice will also call for supplementation of cattle with hay especially for crossbred animals which demand high quality forage to support their higher milk yield potential and hence a need for a cost benefit analysis study to ascertain the feasibility of supplementation. Retention of fewer but more productive crossbreed animals could be an alternative strategy.

 

Conclusion


References
 

Byenkya G S 2004 Impact of undesirable plant communities on the carrying capacity and livestock performance in pastoral systems of South Western Uganda. PhD Thesis, Texas A&M University pp.192 https://txspace.tamu.edu/bitstream/handle/1969.1/255/etd-tamu-2004A-RLEM-Byenkya-1.pdf;jsessionid=DF9C418B42EA9183602041D8EA2B57FC?sequence=1

 

Caltabiano T 2006 Guide to the factors influencing carrying capacities of Queensland’s rangelands. www.nrw.qld.gov.au/factsheets/pdf/land/l80.

 

Coe M J, Cumming D H and Phillipson J 1976 Biomass and production of large African herbivores in relation to rainfall and primary production. Oecologia 22 (4): 341–354

 

Cossins N J and Upton M 1987 cited by De Leeuw P N and Tothill J C 1990 The Borana pastoral system of Southern Ethiopia. Agricultural Systems 25:199 - 218

 

De Leeuw P N and Nyambaka R 1987 The prediction of rangeland production from rainfall data in arid and semi-arid Eastern Africa. In: Dzowela B H. 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. http://www.fao.org/Wairdocs/ILRI/x5491E/x5491e0t.htm

 

De Vos A 1969 Ecological conditions affecting the production of wild herbivorous mammals on grasslands. Advances in Ecological Research (J B Cragg [editor]), Volume. 6, pp: 137-183. London: Academic Press.

 

FAO 1991 Guidelines: land evaluation for extensive grazing. FAO Soils Bulletin 58, FAO, Rome.

 

Grant W E, Pedersen E L and Marín S L 1997 Ecological and natural resources management: Systems analysis and Simulation. John Wiley & Sons, New York. pp. 373

 

Guevara J C, Estevez O R and Torres E R 1996 Utilization of the rain-use efficiency factor for determining potential cattle production in the Mendoza plain, Argentina. Journal of Arid Environments 33: 347 – 353

 

Hocking D and Mattick A 1993 Dynamic carrying capacity analysis as tool for conceptualising and planning range management improvements, with a case study from India.  http://www.odi.org.uk/networks/pdn/papers/34c.pdf

 

Kavana P Y, Kizima J B and Msanga Y N 2005 Evaluation of grazing pattern and sustainability of feed resources in pastoral areas of eastern zone of Tanzania. Livestock Research for Rural Development. Volume 17, Article # 5. Retrieved July 3, 2008, from  http://www.lrrd.org/lrrd17/1/kava17005.htm

 

Kisamba M W, Pender J and Kato E 2006 Impacts of Individualization of Land Tenure on Livestock and Rangeland Management in Southwestern Uganda. Paper presented at the 11th Biennial Conference of the International Association for the Study of Common Property, June 19 - 23, 2006 in Bali, Indonesia. http://www.indiana.edu/~iascp/bali/papers/Kisamba_Mugerwa_W_Pender_Edward.pdf

 

Le Houérou H N, Bingham R L and Skerbek W 1988 Relationship between the variability of primary production and the variability of annual precipitation in world arid lands. Journal of Arid Environments 15: 1 – 18

 

Mei Y, Ellis J E and Epstein H E 2004 Regional Analysis of Climate, Primary Production, and Livestock Density in Inner Mongolia. Journal of Environmental Quality 33:1675 –1681

 

Minson DJ 1981 Effects of chemical and physical composition of herbage eaten upon intake. In: Nutritional Limits to Animal Production from Pastures (editor: J B Hacker). Commonwealth Agricultural Bureaux, Farnham Royal, U.K., pp 169–182

 

Mugerwa J S 1992 Management and utilization of rangeland: The case of Uganda. Paper presented at 2nd  FAO regional workshop on grazing resources for East Africa, Kampala Uganda, 30th  March – 3rd  April 1992.

 

Oesterheld M, Sala O E and McNaughton S J 1992 Effect of animal husbandry on herbivore-carrying capacity at a regional scale. Nature (London) 356: 234 – 236

 

Okello S, Sabiiti E N and Schwartz H J 2005 Factors affecting in sacco dietary degradation by Ankole cattle grazing natural range pastures in Uganda. African Journal of Range and Forage Science 22(3): 157 – 165

 

Phillipson J 1975 Rainfall, primary production and "carrying capacity" of Tsavo National Park, Kenya. East African Wild life Journal 18 (4): 171 - 201

 

Schwartz H J, Walsh M and Harry I 1996 Land use potential in the Ankole Ranching Scheme, Nyabushozi, Uganda. (A report prepared for GTZ, Berlin and College Station)

 

Sserunkuuma D and Olso K 1998 Externalities, risk and the private property-overgrazing paradox: The case of private property farm sin Nyabushozi county, western Uganda.  Center for international food and agricultural policy. http://ageconsearch.umn.edu/bitstream/14372/1/wp98-01.pdf

 

Stella 9.0.2 2007. Technical documentation for Ithink & Stella software. HPS Inc., Hannover, New Hampshire

 

Van Wijngaarden W 1985 Cited by de Leeuw and Tothill JC 1990 Elephants-Trees-Grass-grazers: Relationships between climate, soils, vegetation and large herbivores in a semi- arid ecosystem. IT C publication No. 4, Enschede, Netherlands



Received 16 January 2009; Accepted 6 July 2009; Published 1 September 2009

Go to top