Introduction
The first brief in this series set out definitions of agricultural land and outlined the regulatory framework contained in the Preservation and Development of Agricultural Land Framework Bill. The Bill proposes heavy regulation of agricultural land, which in turn imposes substantial information requirements. Can these requirements be met?
The draft Bill provides for the establishment of an electronic geo-referenced land register which will:
The draft Bill provides for the establishment of an electronic geo-referenced land register which will:
- store data and information for the development, protection, sustainable use and management of natural agricultural resources and agricultural land. This includes demarcation of high potential cropping areas and potential agricultural land. For each piece of agricultural land there must be a record of its ownership, including the nationality and gender of the landowner, and any other information as may be prescribed by the Minister from time to time, and the characteristics of agricultural land, including land cover and land capability class. Information on current agricultural or other land use, environmental encumbrances, water licenses and other natural resource-related information is also required.
- lodge and track applications.
Every provincial Department must provide information on relevant spatial datasets, show the extent of agricultural land lost to mining, formal urban residential developments, informal urban residential developments, and industrial developments, and integrate datasets from different sources, including municipal and farm level.
What resources do we currently have to help meet these requirements?
1. The State Land Audit, completed in 2014, was conducted to determine how much land was owned by the state, what was it used for, and who were the occupants or users. The audit was conducted for all spheres of government, and the former homelands, public land held by the Ingonyama Trust and land of state owned enterprises. It excluded land not registered at the Deeds Office. There were site visits to registered state land, while a desk top study was made of private land.
The audit provided statistical information pertaining to land ownership, specifying gender, and nationality or citizenship, and identity of owner. The audit found that of the 121 973 200 hectares of land in South Africa, 79% was privately owned, 14% was state land and 7% could not be accounted for. Of the private land, 48% was owned by individuals, 22% by companies, 27% by trusts and 3% by private organizations. The 17 061 882 hectares of state land was divided into 1 155 508 land parcels. 40% of this land was held by national departments, including tribal trust land, 22% was held by parastatals, 19% by provinces and 12% by municipalities, with the remainder not classified.
The Surveyor-General also reported to parliament that 95% of unsurveyed land in the former homelands in 2011 had been surveyed by 2013. The Land Audit reported that 16 035 593 hectares of land were situated in the former homelands. Some of this would have been private and some held in trust by the state.
The extent of electronic geo-referencing of land is unclear.
2. On the basis of national spatial data in its possession [1], the Department estimates that the distribution of land capability by class and land by use is:
Class I 0.0% Inland water 0.2%
Class II 1.5% Irrevocably transformed from
Class III 11.5% agricultural use 2.6%
Class IV 13.5% Formally protected (e.g.
Class V 11.2% game reserves) 4.6%
Class VI 14.9% Forestry 1.3%
Class VII 37.3% Cultivation 11.4%
Class VIII 10.3% Range land 80.0%
Again, the extent and precision of geo-referencing is unclear, especially since the class of land can vary over short distances. The 2002 land type data set has a resolution of 25 hectares and there must be close to five million such squares across the country. Each square is classified by majority land use, so all irrevocably transformed areas smaller than 12.5 hectares per square were excluded in the results. Permanently transformed areas smaller than 12.5 hectares per square include features such as roads, rural dwellings, open cast mining, small residential and industrial developments. So the irrevocably transformed area estimate is below what is in fact on the ground. Further studies, the Department concludes, are need to quantify the extent of these areas.
There is another aspect to the information required. Every application for subdivision, rezoning, land consolidation, long leases and acquisition by foreigners will require an agro-ecosystem report. This report must contain the proposed use of the land, municipal and provincial land use frameworks, information on soil, terrain, natural vegetation, climate and water sources, agricultural land capability, on and off farm infrastructure, current agricultural enterprises and uses and an agro-system impact assessment of the change. Only a South African Council for Natural Scientific Professions registered agricultural scientists can do the work.
Done properly, the preparation of such a report will be expensive, to the point of being unsupportable for many small farmers and all micro farmers. If incorporated into the land register these reports will build up more detailed knowledge of agricultural land in a piecemeal fashion and some land will not be the subject of reports for a very long time.
Developing a geo-referenced information system will involve a massive amount of work. In the process, inconsistencies in the Deeds Office records will be uncovered and will have to be dealt with. Unregistered land will have to be dealt with. There is a great deal of potential for land to be assigned to the wrong capability class and extensive applications for changes can be expected. Moreover, there are rights to the occupation of land not recorded in the Deeds Office: the right to occupy land on privately owned farms, on state land and in tribal areas. The Bill envisages that these rights must be registered as servitudes, and that the Minister or MEC must approve them, except in some specified cases.
Can such an elaborate system work at tolerable cost, and with South Africa’s state capacity? The next brief will deal with this issue.
[1] These include:
- National Land capability classification derived from the 1:250 000 land type data set - 2002
- Former homelands and TBVC states – spatial demarcation
- Protected areas - national and provincial as derived by the Department of Environmental Affairs - 2009
- Cultivation – National Field crop boundaries, KZN Landcover 2009 and Inkomati Catchment Management Area Landcover 2010
- Forestry plantations – DAFF 2010; Landcover 2000
- Permanently transformed areas – SPOT Building Count (ESKOM) 2009; Roads 2006; Landcover 2010.
Charles Simkins
Senior Researcher
charles@hsf.org.za
