Thesis (MSc)--Stellenbosch University, 2003. / ENGLISH ABSTRACT: The South African government started the restructuring process of the state’s forest assets in 1998. The
privatisation process includes all the assets of the South African Forestry Company (SAFCOL) and half of the
former homelands’ 150 000 hectares of forest. In August 2000 SAFCOL released their “Operational Plan for
Implementing Exit from Forestry in the Southem-Cape Portion of the Western Cape Region”. This plan
identified only major land uses (agriculture, forestry, and conservation). A more detailed and intensive land
evaluation study was required to specify land utilisation types that are tailor-made to each land unit of the study
area.
The main intention of this research study is to develop a more detailed evaluation process that elaborates on the
land uses proposed by SAFCOL, which is site-specific in terms of the type of agricultural system to be used on
specific areas, or the type of indigenous vegetation to be restored in conservation areas. La Motte plantation was
taken as the case study and the SAFCOL digital database for the study area was used as the input data.
The Automated Land Evaluation System (ALES) was the computer software package used to build the expert
system to evaluate land according to the method presented in the FAO 1976 report. The ALES model built in
this research study had 15 decision trees (one per land utilisation type) resulting in a total of 1678 branches,
which relate land characteristics to severity levels of land qualities. During the computation of an evaluation
ALES attempts to place each map unit into one of the four severity levels of land qualities within each landutilisation
type. Physical suitability of each land unit for each land utilisation type was determined by the
maximum limitation method. ALES is not a GIS and does not by itself display maps. The evaluation result
matrix was exported into ArcMap for further optimisation and geographical analysis to enable the spatial
representation of the results. After completion, taking into account the theoretical background, optimal terrain
units were identified for the different land uses considered and the results are presented as tables and maps.
Fynbos is the most suitable alternative land use for the study area followed by Pears, Sauvignon Blanc and
Chardonnay vines. Pinotage, Shiraz, Cabernet Sauvignon and Cabernet Franc vines were least suitable as
alternatives. The study found that the SAFCOL’s database is not sufficient to meet the requirements of a detailed
site-specific land evaluation process. The polygon attribute table of the soil coverage only provided a subset of
the land characteristics necessary to build and run the model. Data fields like soil form, depth, drainage,
wetness, terrain type, aspect and climatic information had to be created because most of the data provided were
in a non-digital form. The database was not complete and more precise data are needed to improve the system. / AFRIKAANSE OPSOMMING: Die Suid-Afrikaanse regering het in 1998 met die herstruktureringsproses van die bosboubates
van die Staat begin. Die privatiseringsproses het al die bates van die Suid-Afrikaanse
Bosboumaatskappy (SAFCOL) en die helfte van die vorige tuislande se 150 000 hektaar
ingesluit. In Augustus 2000 het SAFCOL sy Operasionale Plan vrygestel vir die
implementering van sy onttrekkingsprogram van bosbou uit die Suid-Kaap gedeelte van die
Weskaap-streek. Hierdie plan het slegs die hoof landgebruike geidentifiseer, bv. landbou,
bosbou en natuurbewaring. ‘n Meer gedetaileerde en intensiewe grondgebruikstudie was
nodig om geskikte gebruikstipes te identifiseer wat optimale altematiewe gebruike
spesifiseer vir elke landeenheid in die studie-area.
Die hoofdoel van hierdie navorsingstudie is om ‘n meer gedetaileerde proses te ontwikkel
ter uitbreiding van die altematiewe landgebruike wat deur SAFCOL voorgestel was.
Hierdie voorstel moet meer ligging-spesifiek wees in terme van die tipe landbougewas of
die tipe inheemse plantegroei wat in natuurbewaringsgebiede gevestig moet word. Die La
Motte-plantasie is as voorbeeld gebruik om hierdie gevalle-studie te doen en die inligting is
vanaf die SAFCOL digitale databasis verkry.
Die rekenaar sagteware-pakket wat gebruik is om die land-evalueringstelsel te bou, is die
“Automated Land Evaluation System” (ALES). Dit berus op die metode wat in die verslag
van die FAO in 1976 voorgestel is. Die ALES model wat in hierdie navorsingstudie benut
is, het 15 beslissingsbome (“decision-trees”) (een per landgebruikstipe) wat ‘n totaal van
1678 vertakkings lewer. Landeienskappe word hierdeur in verband gebring met
verskillende geskiktheidsvlakke vir verskillende gewasse. Gedurende die berekening van
hierdie evaluasie, het ALES elke gebiedseenheid in een van die vier geskiktheidsvlakke per
grondgebruikstipe geplaas. Fisiese geskiktheid van elke landeenheid vir elke
grondgebruikstipe is bepaal deur die maksimum beperkingsmetode. ALES is nie ‘n GIS nie
en op sy eie vertoon dit nie kaarte nie. Die uitslag van die geskiktheidsmatriks is na
ArcMap uitgevoer vir verdere optimisering en geografiese analises ten einde die resultate
ruimtelik voor te stel. Na afhandeling, met inagneming van die teoretiese agtergrond, is optimale terrein-eenhede gei'dentifiseer met inagneming van die verskillende landgebruike
en is die resultate in tabel en kaartvorm aangebied.
Fynbos is die mees geskikte altematiewe landgebruik vir die studiegebied gevolg deur Pere,
Sauvignon Blanc en Chardonnay wingerde. Pinotage, Shiraz, Cabernet Sauvignon en
Cabernet Franc wingerde is minder geskikte altematiewe. Die studie het bevind dat die
SAFCOL databasis nie voldoende was om aan die vereistes van ‘n gedetaileerde
liggingspesifieke landevalueringsproses te voldoen nie. Die poligoon-attribuuttabel van die
grondoorleg het net ‘n subversameling van die landeienskappe verskaf wat benodig was om
die model te bou en uit te voer. Datavelde soos grondvorm, diepte, dreinering, vogtigheid,
terreintipe, hellingrigting en klimaatinligting moes geskep word, omdat meeste van die data
wat verskaf is nie in ‘n digitale vorm beskikbaar was nie. Die databasis was nie volledig nie
en meer presiese data word benodig om die stelsel verder te verbeter.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/53247 |
| Date | 12 1900 |
| Creators | Fernandes Ruiz, Ricardo |
| Contributors | Zietman, H. L., Rozanov, A., Stellenbosch University. Faculty of Arts and Social Sciences. Dept. of Geography and Environmental Studies. |
| Publisher | Stellenbosch : Stellenbosch University |
| Source Sets | South African National ETD Portal |
| Language | en_ZA |
| Detected Language | Unknown |
| Type | Thesis |
| Format | 107 p. : ill. |
| Rights | Stellenbosch University |
Page generated in 0.0032 seconds