Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Afrique du Sud → The ecology of Baobabs (adansonia digitata l.) in relation to sustainable utilization in Northern Venda, South Africa

University of the Witwatersrand (2012)

The ecology of Baobabs (adansonia digitata l.) in relation to sustainable utilization in Northern Venda, South Africa

Venter, Sarah Marguerite

Titre : The ecology of Baobabs (adansonia digitata l.) in relation to sustainable utilization in Northern Venda, South Africa

Auteur : Venter, Sarah Marguerite

Université de soutenance : University of the Witwatersrand

Grade : Doctor of Philosophy (PhD) 2012

The harvesting and commercialization of non-timber forest products (NTFPs) both present social, economic and environmental challenges. Thousands of NTFPs are harvested from wild populations around the world and the use thereof meets the subsistence and commercial needs of many millions of people. Yet few such products are being sustainably harvested, presenting a threat to the survival of many species. As commercial interest in NTFPs grow, it is increasingly important to understand the biology of the species being harvested. A number of species have oil-bearing seeds which provide ingredients for cosmetic formulations. Baobab ((Adansonia digitata L.) seed oil, is one such ingredient, which has rapidly become popular on global markets. Baobabs are an iconic, very long-lived, African savanna tree species known throughout their range as a source of food, fibre and medicine, yet global demand for its fruit derivatives (both seed oil and fruit pulp) have raised concerns about sustainability. This study undertook to investigate the ecological aspects of fruit harvesting and to document the socio-economic benefits of commercializing the fruit. Research was conducted in northern Venda, Limpopo Province, South Africa. Population dynamics, fruit production, phenology and recruitment were investigated in five land-use types, namely : nature reserves, rocky outcrops and plains, representing natural land-use types, and fields and villages representing human-modified land-use types. Density of trees was determined from transects, fruit and flower counts were done on 106 trees over 2-3 seasons, seed viability and seedling/sapling survival rates were determined and interviews with harvesters were conducted in villages across the study area. Data analysis used standard univariate statistical methods and the results were used to create a stage structured population projection matrix model. Villages and fields had higher densities of trees (2.16±0.44 and 1.13±0.52 plants/ha) than plains and rocky outcrops (0.96±0.25 and 0.83±0.24 plants/ha). All land-use types had positively skewed size-class distribution (SCD) curves and negative to flat SCD slopes indicating low recruitment. Quotients and the permutation index suggest that recruitment and mortality are episodic events, but nonetheless that the population has been stable. Mature fruit production was higher in villages and fields (89.59±34.61 ; 88.26±32.20fruit/tree) than nature reserves, plains and rocky-outcrops (1.90±1.15 ; 28.64±12.56 ; 12.56±5.59 fruit/tree). Predation on immature fruit by baboons resulted in 58-85% loss. Inter-annual variation in fruit production was significant, with a 2½-fold difference between the highest and lowest years. Flowering followed a steady-state pattern, lasting 1-5 months and peaking in November. Mean flower numbers/tree (711±72 and 287±33) varied significantly between sequential years, but not fruit-set (average of 20±4%). Baobab seed exhibited high viability (>91%) and formed persistent seed banks. Seed production was substantial (5500±2334 seed/ha) ; thus recruitment does not appear to be seed-limited except in areas where baboons are found. Seedling emergence was staggered over two growing seasons. In plots where livestock were excluded (closed plots) 6.33% of seed emerged with a mean natural germination rate of 328±28 days and of these 94.44% died of moisture stress and insect browsing within 21±5 days. In open plots only 2% of the seeds emerged with a germination rate of 377±5 days and all of them died from livestock browsing within 12±5 days. Planted sapling survival in closed plots (65%) was significantly better than in open plots (10%) where repeated livestock browsing and trampling was responsible for high mortality rates. Matrix model output indicated that high livestock numbers and baboon predation of fruit cause population decline. Under zero to moderate livestock numbers, populations in communal land-use types are able to tolerate fruit harvest rates of between 33-90%. Harvesters were marginalized people to whom the sale of baobab fruit had an income value 4 times higher than its subsistence value. The sale of baobab fruit contributed 38% to the overall cash income received from NTFP sales, helping to alleviate poverty and increase access to a cash economy. It was concluded that the management of baobab populations require an integrated approach incorporating human needs and environmental variables. The long-lived nature of baobab trees provides a buffer against short-term pressures (an excellent example of the ‘storage effect’), but unabated environmental degradation and climate change threaten populations. It is recommended that government agencies, businesses, traditional authorities and harvesters invest in strategies that will protect baobabs and boost recruitment to safeguard future populations and ensure multi-generational benefits for rural people.


Version intégrale (13,3 Mb)

Page publiée le 14 janvier 2019