August 11th, 2014, Published in Articles: PositionIT

by Gakethata Bontsibakae and Emmanuel Tembo, University of Botswana

 

Land, pasture and water are key elements in livestock (beef) business and their utilisation should be managed in a manner that will sustain current and future generations. Demand for land is ever growing and therefore there is a need to find solutions for optimal use of land. One such solution is to use multiple criteria decision analysis (MCDA) to manage this demand for land meant for livestock production.

Livestock boreholes are the highest source of water in the beef industry in Botswana. Botswana’s beef industry is one of the largest contributors to economic sustenance in rural communities. Individual farmers and communities apply to the Land Board for allocation of boreholes. Evidence from this study shows the increasing applications for borehole allocations by individual farmers in Ngamiland. Borehole allocations in the study area have not been done in an optimal way as these are generally allocated based on individual applications with the only criteria being that boreholes should be at least 6 km apart.

Borehole allocation in Botswana is done with guidance of the Tribal Land Grazing Policy which regulates pasture resources. The policy stipulates a distance of 6 km between water points. However for a distance range between 4 km and 6 km, an application is assessed based on the recommendation of a range assessment report from Forestry department. Borehole allocation and rejection is done by Land Boards upon application by individuals. Land Boards are challenged in this process due to numerous factors such as the lack of knowledge about the position of allocated boreholes, the livestock carrying capacity, livestock population and general poor record keeping. In general the whole process of borehole allocation from point identification up to the allocation or rejection of the application is cumbersome and is not carried out considering all possible criteria.

This article presents work done on a study to use geographic information systems (GIS) and multi-criteria data analysis (MCDA) to find suitable borehole points in Ngamiland. GIS and MCDA can be considered to fall under the broad term of land use suitability analysis which seeks to identify the most appropriate spatial pattern for future land use according to specific requirements, preferences or predictors of some activity [1]. Malczewski [2] has cited different areas in which GIS based land use suitability analysis has been applied. These areas include ecological approaches for defining land suitability/habitat for animal and plant species, geological favourability, suitability of land for agricultural activities, landscape evaluation and planning, and environmental impact assessment among others. For example, Jeorin et al. [3] used GIS and outranking multi-criteria analysis to create suitability maps for housing for land-use suitability, Bojorquez-Tapia et al. [4] used GIS to cater for interests of the stakeholders in land suitability assessment.

Background on multi-criteria decision analysis

In order to identify the most appropriate pattern for future land-use, planners are faced with a plethora of competing demands. One way to meet these competing demands is to use a combination of GIS and MCDA. MCDA can be defined as an umbrella term to describe a collection of formal approaches which seek to take explicit account of multiple criteria in helping individuals or groups explore decisions that matter [5]. MCDA is an attractive preposition in natural resource management since it has the ability to mix both qualitative and quantitative data as well as expert opinion on an issue which requires a decision. Belton and Stewart [5] state some of the properties of MCDA as follows:

• It takes account of multiple conflicting criteria
• It helps structure the management problem
• It provides a model that can serve as a focus for discussion
• It offers a process that leads to rational, justifiable, and explainable decisions

Mendoza and Martins (2006) have shown how MCDA has been used in forest and natural resource management having classified the MCDA methods into three broad categories i.e. value measured models, where numerical scores are constructed in order to represent the degree to which one decision option may be preferred to another. Such scores are developed initially for each individual criterion, and are then synthesised in order to effect aggregation into higher level preference models. The second category of models is the goal, aspiration or reference level models in which desirable or satisfactory levels of achievement are established for each criterion. The process then seeks to discover options which are closest to achieving these desirable goals or aspirations. The third type of models are the out ranking models in which alternative courses of action are compared pairwise, initially in terms of each criterion in order to identify the extent to which a preference for one over the other can be asserted.

In aggregating such preference information across all relevant criteria, the model seeks to establish the strength of evidence favouring selection of one alternative over another. What distinguishes these models is that in the case of value measured models the values of alternatives reflect a preference order [5]. In the goal aspiration level model the idea is to come up with outcome scenarios which are expressed in terms of satisfying aspirations or goals for each criterion. The alternatives are then systematically eliminated until a satisfactory level of performance of the criteria has been ensured. The out ranking model focuses on pairwise evaluation of alternatives identifying incomparabilities as well as assessing preferences and indifferences [5].

Study area

The area of study is the Ngamiland District which is managed by Tawana Land Board and is shown in Fig. 1. Ngamiland District is sub-divided into two administration districts; Maun and Okavango, (The headquarters being Gumare 250 km west of Maun). Four local authorities administer the district. These are; the District Administration, Tawana Land Board, Tribal Administration and North West District Council. The district local authorities are coordinated through the District Development Committee, which coordinates and monitors district development and is chaired by the District Commissioner. The District Development Committee is accountable to the Full Council on all matters relating to the development of the district [6]. Ngamiland is further divided into sub-land boards namely; Gumare, Maun, Nokaneng, Sehitlhwa, Seronga and Shakawe sub-land board as shown in Fig. 2.

Methodology

Introduction

In order to carry out the study the data shown in Table 1 was collected. Spatial datasets of existing borehole data in vector format as well as various maps were used to overlay information so as to utilise the criteria of finding the best location of borehole points.

The study involved the use of ArcGIS and other software shown in Table 2 for all spatial datasets and the use of MCDA criteria to determine the optimal location of borehole points.

Spatial Analysis tools in ArcGIS were used to do queries, buffering and overlays. As stated by Jankowski [7] the task of multiple criteria spatial decision analysis needs to be supported by task specific map displays. We found the use of ArcGIS very amenable to this. Informate was used to import livestock borehole points which were later exported into ArcGIS as a table file.

MCDA method

The MCDA method used in this study was the goal aspiration model and was an adaption of the one proposed by Chakhar and Mousseau [8] as described previously. The steps taken in the method were as follows:

• Set the goal/define the problem
• Determine the criteria (factors/constraints)
• Standardise the factors/criterion scores
• Determine the weight of each factor
• Aggregate the criteria
• Validate/verify the result.

Goal setting

The overall goal was to find suitable land for livestock water points for Ngamiland. The goal, objectives and attributes are given in Table 3 and shown in Fig. 4.

Determining factors, constraints and weighting

The criteria and factors that were included in the study are shown in Table 4. The factors were further weighted as shown in Table 5.

Maps were created using reclassification operations and specified cutoffs. In this research, the cutoffs were 6 km for proximity to water points, 2 km (proximity to roads, ranches, fenced wildlife management areas, district boundaries), 4 km and 20 km for proximity to water points, roads, ranches, fenced wildlife management areas, district boundaries, disease control boundaries, and village boundaries respectively. ArcGIS spatial analysis tools were used to do carry out the cutoffs. Then the constraint maps would be combined using conjunctive screen method. The constraint maps were created based on the decision makers’/stakeholders’ preferences, with respect to the relative importance of evaluation based on the cattle predators, cattle competitors, livestock carrying capacity and cattle population. The constraints that were looked at were: cattle predators, cattle competitors, carrying capacity, and cattle population.

• Cattle predators: Cattle predators refer to wild animals which kill cattle. These wild animals include lions, crocodiles, leopards, cheetahs, among others; they have negative impacts on productive cattle farming activities. Cattle predators also pose a big threat to farmers as areas with a high number of wild animals attract fewer number of farmers and less livestock boreholes. The cattle predator data was analysed as follows:
  – Wild animals with high significant values were selected.
  – The total of wild animals was done per potential area division area
  – Then totals were standardised into percentages.

• Cattle competitors: Cattle competitors are animals (both domestic and wild) which compete with cattle for pasture, water and roaming area. They include donkeys, elephants, kudus, zebras and others. They negatively impact on the quality of livestock and increase famers’ running expenses. Farmers who are within the area of high cattle competitors are forced to supplement cattle with feedstock and nutrients, and as such those areas attract lesser farmers and less borehole applications. The cattle competitors’ data was analysed as follows:
  – Cattle competitors with high significant value were selected.
  – Total of cattle competitors was done per potential area division area
  – Then totals were standardised into percentages.

• Carrying capacity: Carrying capacity is described as the average number of animals that a grazing area can be expected to support. It is thus the amount of grazing land which should be made available to an individual animal so that it can be maintained and produce efficiently (milk or beef) without deterioration of the natural resources of that area [9]. The correct number stocking rate was calculated using two factors; the land area and the carrying capacity of such area; and there figures were represented into percentages.
• Cattle population: 2012 Cattle census data was collected from the Department of Veterinary Services. Cattle population is also one of the determining factors for suitability area for boreholes. Crowding of cattle in a small areas leads to poor meat quality, injuries, deforestation, land degradation and water contamination. Therefore, the ideal is to position an adequate proportional number of livestock in a corresponding area, bearing in mind the required minimum distance between boreholes. The cattle potential weights analysis was converted into percentages. The negative values were used to show the area which were heavily overstocked; where the cattle number had exceeded the expected number of cattle per area and 100% values indicated the ones with no livestock.

Results

A stepwise process is depicted to show the results of the exercise which was carried out.

Livestock potential areas

The factors and constraints are combined in a GIS to come up with suitable borehole points. Firstly a map showing livestock borehole potential areas was created; this was after removing unwanted areas and leaving eligible areas where a borehole can be allocated without violating Land Board regulations which require boreholes to be 6 km apart. Fig. 4 shows the result of livestock borehole potential areas. It shows the areas that have the potential to have a borehole and excludes road buffers. Note that Fig. 4 is better read together with Fig. 2 to understand the potential areas. The green colour shows the areas which have livestock potential.

Computation of livestock borehole points

After determining the livestock potential areas the next step was to find how many livestock borehole points can be generated in the potential area. The borehole points were then planned on the potential areas at 6 km apart and calculations were done using Infomate. The computed points were then exported into Excel and imported into ArcGIS and merged with other data layers. A map showing proposed borehole points computed in Infomate was created as shown in Fig. 5.

Standardised weightings of constraints

Standardised weights of cattle predators per area in percentage form were computed for Ngamiland Hunting areas labelled as NG/1, NG/2 etc. in Table 6.

Similar constraint weights were computed for the other constraints identified i.e. the cattle competitors, carrying capacity and cattle population. Livestock potential scenarios were then computed for the areas NG1, NG2 up to NG/37 given the constraints. These potential areas after grouping all the factors are shown in Fig. 6.

Borehole suitability map

All the factors were combined geographically and the decision makers’ judgments were used to determine land suitability. The intensity of all factors was found to be equally important and as such their average was drawn as opposed to a comparison matrix (where evaluation is done based on the priority alternatives). All factors’ percentage weights were considered to determine the suitability of the areas. A borehole suitability map was created which could be used by the Land Board to allocate borehole points. Fig. 7 shows the final suitability map which was generated.

Conclusion

This model provides a very good conceptual framework for deciding where to allocate a borehole and when to allocate based on the identified relevant factors and their standardised weight values. A major constraint factor on this application was un-updated or incomplete data. There is need for different authorities charged with responsibilities of collecting data to have such up to date databases which can be used for analysis. Proper allocation of livestock boreholes can aid in keeping the range plants in optimum condition and would promote soil conservation practice. In order to have such equilibrium, Land Boards should allocate boreholes such that the following are observed:

• Balancing the number of animals with the available forage supply.
• Obtaining uniform distribution of animals over the entire range.

Since different stakeholders are involved in livestock borehole processes, it is necessary for stakeholders to agree on important factors in borehole allocation, how these should be measured, and how they should be combined, paired, ranked, rated or grouped. According to Longley et al [10] an important maxim of MCDA is that it is better for stakeholders to argue in principle about the merits of the different factors and how their impacts should be measured, than to argue in practice about alternative decisions. The combination of MCDA and GIS in determining the optimal borehole position holds much promise if all data is availed to the Land Boards that are the official allocators of boreholes in districts
in Botswana.

Acknowledgement

This paper was published in the AfricaGEO 2014 Proceedings and has been republished here with permission.

References

[1] MG Collins, FR Steiner, and MJ, Rushman: “Land-use suitability analysis in the United States: historical development and promising technological achievements”. Environmental Management 28 (5), pp 611–621, 2001.
[2] J Malczewski: “GIS based Land-use suitability analysis: a critical overview”. Progress in Planning Vol 62, pp3-65, 2004.
[3] F Joerin, M Theriault, and A Musy: ‘Using GIS and outranking multicriteria analysis for land use suitability assessment’. International Journal of Geographical Information Science Vol. 15, No 2. pp 153-174, 2001.
[4] LA Bojorquez-Tapia, S Diaz-Mondragon, and E Ezcurra: ‘GIS-based approach for participatory decision making and land suitability assessment’. International Journal of Geographical Information Science Vol. 15, No 2. pp 129-151, 2001.
[5] S Belton and TS Stewart: Multiple Criteria Decision Analysis. An Integrated Approach. Kluwer Academic Publishers, Massachusetts, 2002.
[6] Ngamiland Development Committee: Ngamiland District Developent Plan 6: 2003 -2009. Botswana Government Printers. Gaborone, 2003.
[7] P Jankowski, N Andrienko, and G Andrienko: ‘Map-centred exploratory approach to multiple criteria spatial decision making’. International Journal of Geographical Information Science Vol. 15, No 2. pp 101-127, 2001.
[8] S Chakhar and V Mousseau: ‘Spatial Multicriteria Decision Making’. In S Shekhar and H Xiong, (eds.), Encyclopedia of Geographical Information Science. Springer, 2007.
[9] DI Field: Basic Ecology for Range Management in Botswana. Ministry of Agriculture, Gaborone, 1978.
[10] PA Longley, MF Goodchild, DJ Maguire, and DW Rhind: Geographic Information Systems and Science.: John Wiley & Sons Ltd, London, 2005.
[11] Tawana Land Board (TLB): Land Allocation Policy. Internal publication, 2010.

Contact Gakethata Bontsibakae, University of Botswana, gake1978@yahoo.com and Emmanuel Tembo, University of Botswana, tembo@mopipi.ub.bw

Related Articles

South African Government COVID-19 Corona Virus Resource Portal

Now Media acquires EngineerIT and Energize from EE Publishers

Latest GIS further supports electric and gas networks

EE Publishers 2020 Year Planners

To readers, customers, suppliers, staff and other stakeholders of PositionIT