Navigation – Plan du site

Subsistence farming and conservation constrains in coastal peat swamp forests of the Kosi Bay Lake system, Maputaland, South Africa

La conciliation difficile entre cultures vivrières et conservation de la nature da les forêts marécageuses côtières de Kosi Bay dans le Maputaland en Afrique du Sud
Retief Grobler, Christoph Moning, Jan Sliva, George Bredenkamp et Piet-Louis Grundling
p. 317-324

Résumés

La conciliation difficile entre cultures vivrières et conservation de la nature dans les forêts marécageuses côtières de Kosi Bay dans le Maputaland en Afrique du Sud

Le Maputaland est un secteur de la province du Kwazulu-Natal, une région située au nord-est de l’Afrique du Sud, près des frontières avec le Swaziland et le Mozambique. Ce secteur correspond d’un point de vue géomorphologique à l’extrémité sud de la plaine côtière du Mozambique et d’un point de vue bio-climatique aux confins méridionaux de la zone tropicale africaine. Aussi de nombreuses espèces tropicales se trouvent ici à la limite de leur aire de répartition.

Le Maputaland accueille par ailleurs l’essentiel des zones tourbeuses d’Afrique du Sud avec 60 % des tourbières nationales. Celles des forêts marécageuses - swamp forests - de la région côtière, bien que les plus riches d’un point de vue biologique, sont parmi les moins bien étudiées alors même qu’elles sont sérieusement menacées de disparition par la très forte pression des communautés agricoles locales.

Les études dont cet article rend compte s’appuient sur une base de données constituée en mai 2003 à partir de 34 sites tourbeux sites au cœur des forêts marécageuses. Elle comporte un ensemble d’informations relatives à la végétation mais également à la pédologie, à l’occupation humaine, aux perturbations anthropiques…. Une classification des tourbières a ensuite été réalisée au moyen du modèle TWINSPAN qui a proposé dix classes, des tourbières intactes à celles largement affectées par les interventions humaines. Une analyse statistique a ensuite permis de mettre en évidence les principaux facteurs de la dégradation des tourbières qui sont liés à la progression des cultures vivrières au cœur même de la forêt. Il apparaît que certaines zones tourbeuses ont désormais un fonctionnement hydrologique gravement affecté, ce qui menace la structure même des forêts marécageuses. A cela s’ajoutent les conséquences sur la qualité de l’eau et la disponibilité en nutriments des eaux lacustres qui, en plus de leurs intérêts écologiques, sont une zone de pêche pour les populations locales.

Cette situation pose un double problème : celui de la conservation des tourbières et des forêts primaires qui les abritent d’une part, celui de l’avenir des populations locales qui peu à peu détruisent les milieux qui les font vivre. Des solutions sont à imaginer pour concilier les intérêts des populations à long terme et le maintien de la biodiversité, cela dans la perspective d’un développement durable.

Haut de page

Texte intégral

1Coastal peat swamp forests (CPSF) can be classified as fresh water forested wetlands (photo 1) that are established on peat soils (photo 3) in open ended riverine interdune valleys in close proximity to a marine water body. Coastal peat swamps forests are highly threatened ecosystems in South Africa, being the second rarest forest type in the whole country and only occur in isolated patches from the Mozambique border to just south of the Msikaba River in the Eastern Cape (Moll, 1980 ; Wessels, 1997). Roughly 3 986 ha of swamp forest occur in Maputaland (fig. 1) of which a great deal is still unclassified due to their remoteness and inaccessibility, the swamp forests on the flat coastal plain of Maputaland form 75 % of all swamp forest found in South Africa which make them very valuable for future conservation (Lubbe, 1997).

Photo 1 : A general view of Kosi Bay coastal peat swamp forest in Maputaland (South Africa)

Photo 1 : A general view of Kosi Bay coastal peat swamp forest in Maputaland (South Africa)

2The largest intact individual peat swamp forest in South Africa occurs adjacent to the Syhadla river, the main source of fresh water for the Kosi Bay Lake System (KLS), with a total area of 880 ha (Wessels, op. cit.) (fig. 1).

Figure 1 : Location of Maputaland and Studied area in South Africa

Figure 1 : Location of Maputaland and Studied area in South Africa

3The Kosi Bay Lake System in the north-eastern corner of Maputaland is a unique chain of 4 linked coastal lakes forming both an estuary and freshwater wetland (Begg, 1980). Due to its uniqueness and subsequent international importance it was classified as a Ramsar site in 1991(Cowan and Van Riet, 1998). CPSF in South Africa and especially those around Kosi Bay are unfortunately under threat from various sources. In the past large areas of swamp forest were cleared for sugarcane farming, forestry, slash and burn agriculture, and eradicating of teste fly and mosquito habitat, but are now provided better protection through legislation and alternative insect pest control practices. Legislation has however been found insufficient in regulating aspects having bearing on communal rural areas, especially relating to wetland cultivation (Kotze, 2002).

4The Tonga people of the Tembe Tribe have been living next to the KLS for centuries and relied heavily on coastal peat swamp forests as a source of fresh drinking water and as the only available productive land for cultivating crops. As the peatlands retain surface water and water released from perched aquifers in the dunes very well and the surface Quaternary dune sediments are nutrient-poor (Grundling, 1996). The inclined swamp forest valleys’ slopes facilitate easy drainage of the peat and resulted in traditional cultivation techniques of cutting drainage ditches and constructing raised beds. This gradient of hydrological modifications enable the gardening of a wide variety of crops, with bananas (Musa xparadisiaca) and madumbes (Colocasia esculenta) being the most common (Grundling, op. cit. ; Kotze et al. 2004).

5Tension between the local Tonga inhabitants and the controlling conservation authority, Ezimvelo Wildlife has been high for many years as they were forcibly removed from their land in 1988 with the proclamation of the Kosi Bay Coastal Forest Reserve. A changing socio-economic environment has placed a greater demand on bananas and other peat swamp forest derived crops, mainly because of the following (Herrington and Compton, 2003) :

6- a general increase in population growth as health services and infrastructure have improved over the last few decades ;

7- many miners have been retrenched as mines closed in Gauteng and the Free State, forcing migrant workers to return home and turn to peat swamp forest gardening as their primary means of survival ;

8- portuguese and Tonga speaking illegal immigrants continue to flock to the area, while the Tembe people do not treat them, especially fellow Tongas, with antagonism, but rather feel obligated to help them, leading to yet more gardens being established ;

9- HIV/Aids is also becoming a serious problem causing yet more poverty as the breadwinners are commonly those most at risk to the disease. As the disease proliferates it creates the need for yet more nutrient supplements to remain productive for as long as possible. The fish kraals at Kosi Mouth and peat swamp forest habitat are the two main reliable natural resources guaranteeing food security.

Methods and materials

10A total of 34 sites surrounding the KLS were visited during May 2003. Individual peat swamp forests sites were affected by a wide array of different land uses as they occurred within and without the boundaries of the Kosi Bay Coastal Forest Reserve. Different peat swamp forest sites were classified into one of four mutually exclusive disturbance categories :

  • - pristine peat swamp forest

  • - recently disturbed peat swamp forest

  • - old disturbed recovering peat swamp forest

  • - active gardening in peat swamp forest.

11Most often the specific category was apparent in the field, but in some cases, especially the old disturbed recovering peat swamp forests, information was gathered from local people that have been farming in the area for sometime. It was however essential that each recorded site had to be a peat swamp forest sometime in the past irrespective of how it looked at the time of investigation. An additional requirement for each sampled site was that it still had to contain a distinguishable peat layer, irrespective of its depth, to ensure that is was still a peat swamp forest or at least had the potential to become one again in the future. Selected peat swamp forest sites were randomly sampled by means of 10x10 m plots, restricted to areas with a relatively homogenous vegetation pattern to ensure an authentic representation of all sampled vegetation types. Inside each plot all the presented plant species were recorded and assigned an appropriate Braun-Blanquet cover-abundance value (Werger, 1974).

12The vegetation in each plot was separately assessed in four different layers : tree (T), shrub (S), herb (H), and liana (L) layer. The height of each layers, when present, was decided on based upon the physiognomy of each plot. If the same species occurred in more than one layer, it was recorded as a separate entity in each one. Data on the habitat were also noted and included a subjective categorical classification of the hydrology of the peat soil surface, as either flooded, wet, moist, or dry. The presence of drainage ditches associated with gardening activities was also noted, as well as peat depth as determined with a standard clay auger up to 1.5 m deep. Additional general notes on the nature of any visual disturbances were also recorded and included remarks on cutting and clearing, fire, and active gardening.

13The importance of local knowledge was highly esteemed during the entire duration of the field work phase and included several informal discussions with farmers on the period of active farming or absence, their main fears and grievances, as well as their attitude towards peat swamp forests and its continued existence.

Photo 2 : Peat soil of the coastal peat swamp forest (Maputaland, South Africa)

Photo 2 : Peat soil of the coastal peat swamp forest (Maputaland, South Africa)

14The collected floristic data were captured into the TURBOVEG data basis (Hennekens, 1996) and thereafter analyzed by means of classification and ordination techniques to identify diagnostic vegetation types. The Two Way Indicator Species Analysis (TWINSPAN) numerical classification procedure (Hill, 1979) was initially used for a classification of the vegetation. This procedure not only distinguishes distinct identified groups, but also the relationships between the groups. The relationship between the different vegetation types are represented in a dendogram. An ordination to identify and quantify the importance of the main environmental factors responsible for the distributing pattern of the plant species in the area was achieved via a Detrended Correspondence Analysis (DECORANA/DCA) (Hill and Gauch, 1980). This ordination procedure was performed by means of PC-ORD4 software and represented in a graph. The ordination differs from the traditional phytosociological usage in that the identified groups are not grouped according to their associated vegetation type, but according to their disturbance category, as indicated above.

Results

15Coastal peat swamp forest vegetation groups are distinguished primarily on the level of human interference, according to the numerical classification (fig. 2). The pristine peat swamp forest group displays a gradient between wet and dry swamp forests. Raphia australis and Syzygium cordatum - Schefflera umbellifera Peat Swamp Forest grow in drier conditions without any manmade modifications to the peat hydrology, while Syzygium cordatum - Voacanga thouarsii Peat Swamp Forest and Syzygium cordatum - Ficus trichopoda Riverine Peat Swamp Forest dominate on progressively wetter valley peatland soils. The dry pristine swamp forest group consists of large endemic stands of Raphia australis palms, distinctive on the western shore of Lake Amanzimnyama, as well as recovered peat swamp forests that have been disturbed long ago and re-established on shallow peat (Syzygium cordatum - Schefflera umbellifera Peat Swamp Forest). Syzygium cordatum exhibits a grade of peat hydrological preferences, but remains absent in the very dry Raphia australis and very wet Ficus trichopoda dominated vegetation groups.

Figure 2 : A dendrogram showing the hierarchy of TWINSPAN divisions for the vegetation data of peat swamp forest of the Kosi Bay Lake System.

Figure 2 : A dendrogram showing the hierarchy of TWINSPAN divisions for the vegetation data of peat swamp forest of the Kosi Bay Lake System.

16The heavily disturbed peat swamp forest vegetation group is characterized by a general lack of a fully developed tree layer as well as hydrological modifications made to the peat by means of drain ditches (fig. 2 and fig. 3). Active gardening related peat swamp forest destruction (photo 3A and B) mainly consists of bananas (Musa xparadisiaca) cultivated on the drier upper swamp forest peatland soils (Musa xparadisiaca Garden) (photo 3C) and madumbes (Colocasia esculenta) (photo 3D) grown in both wet and dry conditions normally towards and around the centre of the valley peatland (Colocasia esculenta Garden). Bananas are more dependant on sunlight, which requires a more open canopy and leads to greater clearing and cutting of swamp forest trees (photo 4). Drier wetland conditions preferable for banana growth is created by drainage ditches running perpendicular to the valley’s length, cutting through the peat and causing intensified peat decomposition. Site investigation and dialogue identified bananas as the most favoured crop species by swamp forest farmers where growing conditions are favourable or can be modified to be so. Banana gardening causes therefore more peat swamp forest destruction than any other crop type. Madumbes also known as taro or cocoyams cause the least impact on peat swamp forest exacerbation as it grows well in wet or dry and shaded or unshaded conditions (De La Pena, 1983). The minimal human interference required for its cultivation, in terms of alterations to the existing peat and vegetation cover, merits its use as the most preferred crop species from a conservational perspective. As gardening drives peat decomposition and the peat becomes mixed with the underlying sand the fertility and water holding capacity is progressively lost from the edge of the valley peatland inwards. The result for the farmers is a succession of crops from bananas, madumbes and sugarcane (Saccharum officinarum) to smaller and less nutrient rich vegetables such as, sweet potatoes (Ipomoea batatas), tomatoes (Lycopersicon esculentum), and pumpkin (Cucurbita pepo) before gardens are finally abandoned.

Photo 3 : A, B, C and D : Gardening inside the Kosi Bay coastal peat swamp forest (Maputaland, South Africa). Bananas (Muxa xparadisiaca) and Madumbes (Colocasia esculenta) were planted after the clearing of the peat swamp forest

Photo 3 : A, B, C and D : Gardening inside the Kosi Bay coastal peat swamp forest (Maputaland, South Africa). Bananas (Muxa xparadisiaca) and Madumbes (Colocasia esculenta) were planted after the clearing of the peat swamp forest

Figure 3 : A scatter diagram of DECORANA ordination illustrating the correlation of identified disturbance categories with the main environmental factors.

Figure 3 : A scatter diagram of DECORANA ordination illustrating the correlation of identified disturbance categories with the main environmental factors.

17Disturbed peat swamp forest vegetation units in which no cultivation is currently taking place could be caused by a small number of factors : fire and natural die off are prone to Raphia australis - Cladium mariscus Disturbed Peat Swamp Forests where the drier conditions place the palm dominated forest with abundant dry biomass at risk to fire. Natural gaps are created in the vegetation strata as cohorts of monocarpic Raphia australis palms die and fall over after flowering. Raphia australis is endemic to the area and therefore well protected (Van Wyk, 1994) with anthropological impacts limited to the collection of fallen rachi for construction material (Lubbe, op. cit.). Traditionally this light and durable wood has been a very important component in Tonga livelihood, but one that is increasingly viewed as a sign of poverty and therefore only utilized by those that have no other means available (Ngubane, pers. comm.). Wetter areas where cultivation has ceased in the not too distant past remain as open swamps, typically at valley bottoms where conditions have become too wet and draining insufficient to sustain farming. These open swamps are characterized by vegetation groups such as Typha capensis - Cyperus prolifer Open Wetland and Cyperus textilis - Ipomoea batatas Relic Garden. The latter still contains some sweet potatoes individuals, as well as shaped mounds traditionally used to cultivate smaller vegetables species in wet conditions and remain as "fossil gardens" (Grundling, op. cit. and pers. comm).

Photo 4 : The landscape of the Kosi Bay coastal forest after several years of gardening by the Tonga people of the Tembe tribe

Photo 4 : The landscape of the Kosi Bay coastal forest after several years of gardening by the Tonga people of the Tembe tribe
  • 1. Pristine Raphia australis Peat Swamp Forest

  • 2. Syzygium cordatum - Bridelia micranta Peat Swamp Forest

  • 2.1 Syzygium cordatum - Schefflera umbellifera Peat Swamp Forest

  • 2.2 Syzygium cordatum - Voacanga thouarsii Peat Swamp Forest

  • 2.3 Syzygium cordatum - Ficus trichopoda Riverine Peat Swamp Forest

  • 3. Colocasia esculenta Garden

  • 3.1 Colocasia esculenta - Lycopersicon esculentum Garden

  • 3.2 Colocasia esculenta - Erythrina lysistemon Garden

  • 4. Musa xparadisiaca Garden

  • 5. Raphia australis - Cladium mariscus Disturbed Peat Swamp Forests

  • 6. Typha capensis - Cyperus prolifer Open Wetland

  • 7. Cyperus textilis - Ipomoea batatas Relic Garden

18The importance of anthropological factors in determining the distribution of different distur-bance classes within peat swamp forests is evident from the DCA ordination. Measurable attributes best correlated with Axis 1 (eigenvalue = 0,729, length of gradient = 4,488) include the percentage tree canopy cover (tree co, r-value = -0,81), presence of drain ditches (abbreviation : drain di ; r-value = 0,691) and pristine relevés (abbreviation : pristine ; r-value = -0,472) (fig. 3). These features highlight the distinct difference in the structural intactness of the various disturbance categories, as displayed by the non-continuous relevé distribution paralleled to Axis 1. An intact tree canopy is prevalent among the Pristine and Old disturbed recovering groups, while Gardening and Recently disturbed groups generally lack an integral canopy, except where gardening occurs underneath tall trees, as at the Colocasia esculenta - Lycopersicon esculentum and Colocasia esculenta - Erythrina lysistemon gardens of relevés SYD-1 and MAL-1 respectively. Or the recent disturbance has been limited to the clearing and burning of the herb and shrub layer in preparation or extension of gardening practices, as at the Syzygium cordatum - Bridelia micranta Peat Swamp Forest and Colocasia esculenta - Lycopersicon esculentum Garden vegetation groups of relevés SY-2, NK-1 and SF1-2 respectively. Pristine and Old disturbed recovering groups exhibit little difference in their tree canopy variation as a measure of structural intactness.

19The highest correlated attribute with Axis 2 (eigenvalue = 0,564 ; length of gradient = 4,981) in the vertical direction is the hydrology or wetness of peat soils (abbreviation: hydr ; r-value = 0,611) (fig. 3). Pristine and Old disturbed recovering groups are of an intermediate wetness as the drier Raphia australis Peat Swamp Forest relevés of RA-1 and RA-2 are the only real exceptions. The defi-nite disjunction between Gardening and Recently disturbed identified groups along Axis 2 is a clear indication in the different hydrological regimes between the two. The Recently disturbed group is commonly cleared of vegetation, including the tree layer, but has not yet been drained as gardening still has to commence and remains therefore wetter as opposed to the Gardening group. Intermediate Gardening group relevés such as SYD-1 and MAL-1 belong to the Colocasia esculenta Garden vegetation group, which grow well in a wide variety of hydrological conditions (see text above), while some farmers continue to grow bananas in spite of wet conditions as at relevé NK-3 (Musa xparadisiaca Garden).

Conclusion

20Prior to the interference of man on the intensity level of the last few decades, peat swamp forests in the Kosi Bay Lake System remained in a near to pristine-like condition with a characteristic forests vegetation and woody peat composition (Grundling et al., 1998 ; Grundling et al., 2000). Without the external influences of humans, especially effective in altering the vegetation’s structure and peat moisture contend, the ecosystem’s functioning is principally depended upon their wet hydrological regime. As hydrology was the main factor that best explained the relationship between the individual relevés of the least modified Pristine and Old disturbed recovering groups (fig. 3).

21Traditional Tonga practices enabled the community to live in harmony with their wetlands for several generations, as they rotated their gardens on a chosen peatland (Grundling, op. cit.) and still have wetland habitat that remain intact in spite of their presence over several centuries. Population densities have, however, increased since, bringing these systems under increased pressure, which impede fallowing and lead to progressive peat decomposition as abandoned agriculture land becomes occupied by others. This stresses the need to introduce and implement best management practices pertaining to peat swamp forest gardening, such as minimizing drainage, building awareness and utilizing compost as a growth medium supplement to peat (Kotze et al., op. cit.).

22Tourism in this largely pristine area is playing an increasingly important role in the local economy as access and tourist numbers improve due to better infrastructure, such as new roads and lodges, and wider circulated publicity. It is clear that the local inhabitants know how to manage the wetlands for their own benefit, reconciliation between their traditional practices and Western-style conservation could really boost ecotourism and the local economy (Grundling, op. cit.). Up to now tourism has been unsuccessful to bring about the desired social and environmental benefits to the Kosi Bay area, and while tourism has the capability to make a definite contribution to the situation it needs to be administrated with care (Kotze et al., op. cit.). The different perspec-tives of conservation and utilisation become incompatible in the sense that the age old inherent practice of sustainable wetland utilisation is threatened : the demand on environmentally derived commodities have drastically increase in order to attain financial freedom in an escalating capitalistic driven society, which have caught up with the people from Kosi Bay. The direct result is the shift of peat swamp forest gardening from simply a subsistence activity to a commercial farming pursuit. The latter practice not being justifiable on such a threatened and important functioning ecosystem. Having said this, swamp forest peat soils still provide the only truly viable habitat for cultivation, due to its water retainability, reasonable fertility and ease of draining. It therefore continues to be the agricultural refuge for many remaining subsistence level farmers in the otherwise marginal landscape.

23How to find the best balance between the different peat swamp forest stances remain an unsolved challenge, which needs to address the prime agents responsible for peat swamp forest’s decline in the Kosi Bay Lake System. Answers point beyond the mere exclusion of slash and burn farmers, but calls for the building of trust and mutual understanding between the concerned stakeholders. With this foundation in place the intricate and multifaceted pull of poverty, a key superseding driving force in habitat destruction and unsustainable livelihood, could be contested.

Haut de page

Bibliographie

BEGG G.W., 1980, The Kosi system : aspects of its biology, management and research, in Burton M.N. and Cooper K.H. (eds), Studies on the Ecology of Maputaland, Rhodes University and Wildlife,p. 358-373.

COWAN G.I. and VAN RIET, W., 1998, A Directory of South African Wetlands, Research Report Series, South African Wetlands Conservation Program. Department of Environmental Affairs et Tourism, Pretoria (unpublished).

DE LA PENA R.S., 1983. Agronomy, in Wang J.K. (eds), Taro, a review of Colocasia esculenta and its potentials, University of Hawaii Press, Honolulu,p. 160-179.

GRUNDLING P-L., 1996, Sustainable Utilisation of Peat in Maputaland, KwaZulu-Natal, South Africa, in Abstracts of the 10th International Peat Society Congress, Lüttig, G.W. editions, Bremen, 6 p.

GRUNDLING P.-L., MAZUS H. and BAARTMAN L., 1998, Peat resources in northern KwaZulu-Natal wetlands : Maputaland, Research Report Series, South African Wetlands Conservation Programme, Department of Environmental Affairs and Tourism, Pretoria (unpublished).

GRUNDLING P.-L., BAARTMAN L., MAZUS H. and BLACKMORE A., 2000, Peat resources of KwaZulu-Natal wetlands : Southern Maputaland and the North and South Coast, Council for Geoscience, Report no: 2000-0132 (unpublished).

HENNEKENS S.M. 1996, TURBO(VEG) : software package for input, processing, and presentation of phytosociological data. User’s guide, Version July 1996, Wageningen, IBN-DLO 5 (no pagination).

HERRINGTON S and COMPTOM R., 2003, Greater St Lucia Wetland Park - Maputaland conflict, SABC2, 50/50, March 23 (unpublished).

HILL M.O. 1979, TWINSPAN – A FORTRAN program for arranging multivariate data in a ordered two-way table by classification of individuals and attributes, Ithaca, New York, Cornell University (no pagination).

HILL M.O. and GAUCH H.G. 1980, Detrended correspondence analysis : an improved ordination technique, Vegetatio, 42, p.47-58.

KOTZE D.C., 2002, Wetland cultivation and the rural poor in South Africa : reconciling conflicting needs, Food and Agriculture, Organization of the United Nations, Sub-Regional Office for East and Southern Africa, Harare (unpublished).

KOTZE D.C., GRUNDLING P.-L. and SILVA J., 2004, Socio-economic impacts on the tropical peat swamp forests of Maputaland, Southern Africa (Unpublished).

LUBBE R.A, 1997, Vegetation and flora of the Kosi Bay Coastal Forest Reserve in Maputaland, northen KwaZulu-Natal, South Africa, M.Sc. dissertation, University of Pretoria, Pretoria (no pagination).

MOLL E.J., 1980, Terrestrial plant ecology, in Burton M.N. and Cooper K.H. (eds), Studies on the ecology of Maputaland, Rhodes University and Wildlife,p. 52-68.

VAN WYK A.E., 1994, Maputaland-Pondoland Region, in Davis S.D., Heywood V.H. and A.C. Hamilton (eds), Centers of plant diversity. A guide and strategy for their conservation, Volume 1, IUCN Publications Unit, Cambridge,p. 227-235.

WERGER M.J.A., 1974, On the concepts and techniques applied in the Zürich-Montpellier method of vegetation survey, Bothalia 11(3),p. 309-332.

WESSELS N.G., 1997, Aspects of the ecology and conservation of swamp forests in South Africa, Unpublished M.Tech thesis, Port Elizabeth Technikon, Port Elizabeth, 155 p.

Haut de page

Table des illustrations

Titre Photo 1 : A general view of Kosi Bay coastal peat swamp forest in Maputaland (South Africa)
URL http://geocarrefour.revues.org/docannexe/image/842/img-1.png
Fichier image/png, 844k
Titre Figure 1 : Location of Maputaland and Studied area in South Africa
URL http://geocarrefour.revues.org/docannexe/image/842/img-2.png
Fichier image/png, 111k
Titre Photo 2 : Peat soil of the coastal peat swamp forest (Maputaland, South Africa)
URL http://geocarrefour.revues.org/docannexe/image/842/img-3.png
Fichier image/png, 474k
Titre Figure 2 : A dendrogram showing the hierarchy of TWINSPAN divisions for the vegetation data of peat swamp forest of the Kosi Bay Lake System.
URL http://geocarrefour.revues.org/docannexe/image/842/img-4.png
Fichier image/png, 27k
Titre Photo 3 : A, B, C and D : Gardening inside the Kosi Bay coastal peat swamp forest (Maputaland, South Africa). Bananas (Muxa xparadisiaca) and Madumbes (Colocasia esculenta) were planted after the clearing of the peat swamp forest
URL http://geocarrefour.revues.org/docannexe/image/842/img-5.png
Fichier image/png, 652k
URL http://geocarrefour.revues.org/docannexe/image/842/img-6.png
Fichier image/png, 615k
URL http://geocarrefour.revues.org/docannexe/image/842/img-7.png
Fichier image/png, 523k
URL http://geocarrefour.revues.org/docannexe/image/842/img-8.png
Fichier image/png, 657k
Titre Figure 3 : A scatter diagram of DECORANA ordination illustrating the correlation of identified disturbance categories with the main environmental factors.
URL http://geocarrefour.revues.org/docannexe/image/842/img-9.png
Fichier image/png, 51k
Titre Photo 4 : The landscape of the Kosi Bay coastal forest after several years of gardening by the Tonga people of the Tembe tribe
URL http://geocarrefour.revues.org/docannexe/image/842/img-10.png
Fichier image/png, 438k
Haut de page

Pour citer cet article

Référence papier

Retief Grobler, Christoph Moning, Jan Sliva, George Bredenkamp et Piet-Louis Grundling, « Subsistence farming and conservation constrains in coastal peat swamp forests of the Kosi Bay Lake system, Maputaland, South Africa », Géocarrefour, Vol. 79/4 | 2004, 317-324.

Référence électronique

Retief Grobler, Christoph Moning, Jan Sliva, George Bredenkamp et Piet-Louis Grundling, « Subsistence farming and conservation constrains in coastal peat swamp forests of the Kosi Bay Lake system, Maputaland, South Africa », Géocarrefour [En ligne], Vol. 79/4 | 2004, mis en ligne le 31 mai 2008, consulté le 23 octobre 2017. URL : http://geocarrefour.revues.org/842 ; DOI : 10.4000/geocarrefour.842

Haut de page

Auteurs

Retief Grobler

University of Pretoria
Department of Botany
Hatfield 0002
South Africa
E.mail : retief@tuks.co.za

Christoph Moning

Technische Universitate Munchen
Chair of Vegetation Ecology 85350 Freising
Germany

Jan Sliva

Technische Universitate Munchen
Chair of Vegetation Ecology 85350 Freising
Germany

George Bredenkamp

University of Pretoria Department of Botany
Hatfield 0002
South Africa

Piet-Louis Grundling

Working for Wetlands Programme
P.O. Box 912924
Silverton 0127
South Africa

Haut de page

Droits d’auteur

© Géocarrefour

Haut de page
  • Revues.org