Factors affecting the conservation of renosterveld by private landowners

McDowell, Clive Robert

Factors affecting the conservation of renosterveld by private landowners

Doctoral Thesis

1988

Publisher

University of Cape Town

Department

Department of Environmental and Geographical Science

Faculty

Faculty of Science

Abstract

West Coast Renosterveld, once prevalent on the south-western Cape lowlands of the floristically rich Cape Floral Kingdom, is now South Africa's scarcest vegetation type. Delimitation and measurement of the 55 "island-remnants", scattered amongst agricultural lands indicated that only 3% of the original vegetation remains. Measured soil quality, gradient and rainfall (critical agricultural criteria) were used to derive agricultural 'threat' indices (probability of agricultural clearance) for each remnant. The indices agreed well with observed trends in the recent clearance of remnants not deliberately conserved by landowners. This new approach has potential for determining which remaining natural sites most urgently need preservation. It was demonstrated that agro-technical innovation poses a long-term threat to even the low agricultural 'threat' rated renosterveld. Pasturage, invariably practised in renosterveld, was assessed by comparing "grazed" with "ungrazed" plant species covers. Within the test site, heavy grazing increased Asteraceae and Iridaceae, decreased Poaceae and Rutaceae, and eradicated Proteaceae. These changes may reflect local trends associated with modern grazing regimes. Total plant diversity and cover were not found to be affected. Therefore, controlled pasturage reflects a relatively minor threat to the extinction of floristic elements. Private landowners are found to control the destiny of 80% of West Coast Renosterveld. The conservation attitudes and behaviour of a random sample of these critical decision-makers were analysed. A new approach was devised to assess 32 subjective, bias-prone "intangible" variables. This required additional assessors to provide independent, non-parametric ratings of the author's tape-recorded interviews with the landowners. Results were pooled and "inter-assessor" measurement error was estimated. Altogether 52 variables, including a further 20 "tangible" variables, were rated under categories: 'Demographic', 'Psycho-Social, 'Land Use' and 'Conservation Strategy'. A correlation matrix portrays inter-relationships between variables and their correlations with landowner "Conservation Behaviour". Landowners' knowledge of biota, rapport with interviewer, education, affluence, bilingualism, and parents' education all correlated positively with conservation. Linear equation models were derived (using Best Subsets Statistical Programme) to predict 'Conservation Behaviour'. Although most conservation related variables are "fixed" (e.g. it is not possible to change parents' education), the understanding of these background factors (not previously applied to conservation of natural ecosystems) enables choice of the most appropriate strategy to persuade landowners to conserve. Different groupings of landowners having similar ratings of key variables have similar needs and constraints to be considered vis a vis improving their "Conservation Behaviour".