The herbaceous vegetation of subtropical freshwater ...

South African Journal of Botany 104 (2016) 158–166

Contents lists available at ScienceDirect

South African Journal of Botany journal homepage: www.elsevier.com/locate/sajb

The herbaceous vegetation of subtropical freshwater wetlands in South Africa: Classification, description and explanatory environmental factors E.J.J. Sieben a,⁎, T. Nyambeni b, H. Mtshali b, F.T.J. Corry d, C.E. Venter e, D.R. MacKenzie f, T.E. Matela b, L. Pretorius g, D.C. Kotze c a

Department of Plant Sciences, University of the Free State, Qwaqwa Campus, Private Bag X13, Phuthaditjhaba 9866, South Africa Department of Plant Sciences, University of the Free State, South Africa Center for Environment and Development, University of KwaZulu-Natal, South Africa d Freshwater Research Unit, Department of Zoology, University of Cape Town, South Africa e Kyllinga Consultants, Pretoria, South Africa f Plant Specialist Group Mpumalanga, Lydenburg, South Africa g Applied Behavioral Ecology Research Unit, Department of Environmental Sciences, University of South Africa, South Africa b c

a r t i c l e

i n f o

Article history: Received 8 February 2015 Received in revised form 30 October 2015 Accepted 30 November 2015 Available online 5 February 2016 Edited by JB Adams Keywords: Phytosociology Peatland Helophytes Cluster analysis Ordination

a b s t r a c t In this paper, the wetland vegetation types of subtropical wetlands occurring in South Africa are discussed. The South African National Wetland Vegetation Database targeted the collation of all available wetland vegetation data consisting of species composition, Braun-Blanquet cover-abundance data and the relevant environmental parameters. A subset of this database that represented subtropical wetlands, was used for analysis by means of clustering and ordination techniques. Forty-nine wetland communities are described and these are summarized into sixteen community groups. The most important factors that account for the variation among these wetlands are the soil clay percentage and associated soil electrical conductivity. Some communities are associated with peaty or sandy soils with a very low soil electrical conductivity on the coastal plain while others that occur more in loamy soils with a high soil electrical conductivity are associated with drainage channels, at the foothills of mountains or the escarpment. A specific group of wetlands is associated with nutrient-poor substrates of the Msikaba group Sandstones in Pondoland or on coastal sands of Maputaland. Subtropical wetlands are particularly vulnerable to degradation due to cultivation as they are found in some of the most densely inhabited rural areas of South Africa. © 2016 SAAB. Published by Elsevier B.V. All rights reserved.

Introduction In their comprehensive vegetation atlas of South Africa, Mucina and Rutherford (2006) note that they were the first authors to include azonal vegetation such as wetlands into the vegetation map of South Africa. Historically, vegetation mapping in South Africa has focused on terrestrial vegetation, and its management aspects were informed by correlation of vegetation with climatic factors or large-scale environmental impacts such as grazing and fire (Rutherford and Westfall, 1986; Tainton, 1999). Wetlands often form only small portions of a landscape, and if they already were included in large-scale vegetation studies they were rarely regarded as separate entities. Recently, however, it has been recognized that wetlands play a disproportionate role in large-scale ecological processes (Mitsch and Gosselink, 2000; Keddy, 2004;

⁎ Corresponding author at: Department of Botany, University of the Free State, Qwaqwa Campus, Private Bag X13, Phuthaditjhaba 9866, South Africa. E-mail address: [email protected] (E.J.J. Sieben).

http://dx.doi.org/10.1016/j.sajb.2015.11.005 0254-6299/© 2016 SAAB. Published by Elsevier B.V. All rights reserved.

Schlesinger and Bernhardt, 2013) and they provide many ecosystem services important to rural development (Kotze et al., 2008). In the last two decades, wetlands have become some of the most intensively managed parts of the rural landscape in South Africa and they feature as important elements in strategic plans for water resource management (Driver et al., 2011; Nel et al., 2011). They also represent heavily utilized parts of the landscape in rural areas where people are directly dependent on the resources and services that nature provides. Therefore, the wise use of wetland resources is often promoted and it is certainly desirable to obtain more information about their species composition, ecology and distribution. Wetland vegetation differs significantly from upland vegetation, as plants in areas that are inundated for at least part of the year have to cope with very specific stresses (Cronk and Fennessey, 2001; Keddy, 2004). Inundation of soils creates anaerobic conditions in the soils which lowers the redox potential and drastically changes the biochemistry of soils (Schlesinger and Bernhardt, 2013). Many reduced compounds in such soils are toxic to plant growth and plants require specific physiological adaptations to deal with them. But most of all,

E.J.J. Sieben et al. / South African Journal of Botany 104 (2016) 158–166

plants need adaptations to manage their own energy needs in the root zone. Respiration in the root zone requires oxygen and therefore oxygen is transported from the surface to below the soils in specialized tissues called aerenchyma (Cronk and Fennessey, 2001). Different habitats in the wetland have different inundation regimes during the course of a year and in most cases, wetlands can be subdivided into habitats that are temporarily, seasonally or permanently wet. These habitats can be recognized by means of their soil hydromorphic features (Kotze et al., 1996). Plants are often adapted to germinate in periods when the water level drops (Shipley et al., 1989) but they may form extensive mats by the time the water level rises again. Many wetland plant communities are dominated by clonal dominant matrix species with interstitial species in between (Boutin and Keddy, 1993). For this reason, it is generally easy to recognize and classify wetland plant communities as most communities have a distinct dominant species. Wetlands of warm tropical and subtropical areas are among the most productive ecosystems in the world and most African wetlands fall into this category (Thompson and Hamilton, 1983). In South Africa, such wetland ecosystems are only present in the northern provinces and along the coast at the lower altitudes. This large number of wetlands contains a wide variety of vegetation types linked to specific environmental conditions such as soil type, wetness and nutrient contents. In this paper, we provide an overview of vegetation types of subtropical wetlands in South Africa, following the definition of Azf6 (subtropical freshwater wetlands) by Mucina and Rutherford's (2006). This includes those wetlands in the northern and eastern provinces of South Africa between altitudes of 0 and 1400 m, that are covered by medium short to tall reeds, sedges or rushes. Many of the widespread wetland vegetation types, such as those dominated by Kweek (Cynodon dactylon), Common reed (Phragmites australis), Imperata cylindrica or Leersia hexandra are also found in subtropical wetlands, either extensively or in small patches, but they have not been dealt with in the current paper. These vegetation types and their connection to subtropical wetland vegetation are discussed in the report by Sieben et al. (2014). The aims of this study are to present a classification based on vegetation composition, that is useful for conservation planning and management, as well as to understand and differentiate the communities on the basis of environmental factors that determine the distribution and abundance of the plant species that make up these communities. Methods The standard method for collecting vegetation data in South Africa is the Braun-Blanquet method (Westhoff and Van der Maarel, 1978;

159

Brown et al., 2013). This involves sampling vegetation in representative quadrats in terms of species composition and each species' relative abundance. These data can then be entered into a database programme such as Turboveg (Hennekens and Schaminée, 2001), where it can be made available for different kinds of analysis. The data that were already available in this format was brought together and combined with new data from areas that were under-represented in historical wetland vegetation studies, which was nearly everywhere in the subtropical areas except for Maputaland. The overall effort to combine wetland vegetation data from historical studies and supplement it with new data gave rise to the South African National Wetland Vegetation Database (Sieben et al., 2014). Subtropical wetland vegetation is represented by 1112 vegetation plots, of which 388 plots were newly collected. The source of the other data, from historical studies, is indicated in Table 1. The datasets indicated in grey contribute only a very small number of plots to the overall dataset of the current study. For the new data that was collected, a standardized sampling protocol was developed so that the Braun-Blanquet vegetation data could be supplemented by standardized environmental data, which has relevance for wetland habitats (Table 2). Additionally, in at least one plot per wetland for the new data, a soil sample from the topsoil (the rooting zone of the vegetation) was collected and analysed to provide additional explanatory variables. The environmental data that were used for the analysis are summarized in Table 2, together with their methods of measurements. The data that were not available for all vegetation plots are indicated with an asterisk. In those vegetation plots where no soil data was collected, some general environmental characteristics were assessed on site. Analyses that required the most detailed explanatory data (ordination analysis) were restricted to a subset of the total database where all data fields were complete. Two important environmental characteristics listed in Table 2 are the Hydrogeomorphic Unit and the Hydroperiod. The Hydrogeomorphic (HGM) Unit refers to the position of the wetland in the landscape and indicates the reason why that position in the landscape has a net surplus of water (for at least part of the year) or an impeded drainage. This kind of wetland classification has been discussed in detail by Ollis et al. (2013). Within the current study, all sites have been classified either as Floodplain, Depression, Seepage, Valley bottom or Channelled Valley Bottom wetlands, depending on the main types of water inflows and outflows that occur in these wetlands. The second most important environmental character is the Hydroperiod, which is the time period of the year in which the area is inundated or saturated. Unfortunately, it is not possible to obtain detailed measured data on this variable on a large scale; it has to be inferred from the soil hydromorphic features, following Kotze et al. (1996).

Table 1 Historical studies that were included in the compilation of a database for subtropical wetland vegetation in South Africa.

Publication/dataset

Area

Comments

Furness (1981) Schoultz (2000) Taylor (2000) Neal (2001) Goge (2002) Venter (2002) Sieben et al. (2006) Grobler (2009) Corry (2011) Pretorius (2011) Sieben et al. (2014) Coetzee, Bredenkamp & Van Rooyen (1994) Perkins et al. (2000) Kareko (2002) Sieben, Kotze & Morris (2010) Cowden, Ellery, Kotze, & Sieben (2014) Collins (2011)

Pongola floodplain Mkhuze Swamps Mdlanzi Pan, Zululand Mkhuze floodplains Eastern shores Lake St. Lucia Mfabeni Swamp, Eastern shores Wetlands South of Richards Bay Swamp Forest Kosi Bay Western Cape lowlands Wetlands Nothern Maputaland Catalina Bay, KwaZulu-Natal Wetlands of Heidelberg, Witbank, Pretoria area Southern KwaZulu-Natal Middelvlei Wetland, Stellenbosch Maloti-Drakensberg Transfrontier Park Rehabilitated wetlands in KwaZulu-Natal Free State pans and Valley bottoms

Vegetation plots 10 × 10 m Plots on transects, no wetness data Plots on transects, no wetness data Plots on transects, no wetness data One large wetland sampled Unpublished consultancy Limited number of sedgeland plots, various stages of succession Soil data and many other variables available for most plots Masters thesis, soil data not yet available Vegetation sampled twice in permanent quadrats

No. of plots 17 73 51 127 130 189 11 5 42 46 25 2 1 1 2 1 1

160


Table 2 Environmental variables that were included in the analysis for environmental comparison of the different plant communities in subtropical wetlands. Nominal variables are variables that consist of several categories.

Variable

Type of variable

HGM_type

Nominal

Soil depth

Nominal

Wetness

Index

Soil texture (1)

Nominal

Soil texture (2)

Ratio*

Soil organic matter (1)

Index

Soil organic matter (2) pH Inundation Altitude Slope Electrical conductivity Nitrogen

Ratio*

Walkley-Black method

Ordinal* Ratio Ratio Ratio Rato*

Water extraction of soil, using pH meter.

Phosphorus Major cations

Ratio* Ratio*

Ratio*

Method of measurement

Abbreviations or categories used in ordination diagram

Units

Level 4 of SAWCS classification system (Ollis et al., 2013), Depression, Floodplain, Valley bottom, Valley bottom with channel, Seepage, Channel Soil augering, subdividing into two categories: deep soils (N50 cm), shallow soils (b50 cm) Assessment of soil hydromorphic features following Kotze et al. (1996). Index: 1 = no wetland, 2 = temporary, 3 = temporary/seasonal, 4 = seasonal, 5 = semi-permanent, 6 = permanent Feel of soil by touch

n.a.

Depression, Fldplain, VB, VB-wc, Seepage, Channel

n.a.

Soil_dp, Soil_sh

n.a.

Wetness

n.a.

Peat, Sand, Loam, Clay, Silt, Gravel %Sand, %Clay, %Silt

mass %

Sieving and subdividing into three fractions Clay (b0.002 mm), Silt (0.002–0.05 mm), Sand (0.05–2 mm) Checking colour of soil: mineral soils = 1, dark or humic soils = 2, peaty soils = 3

using GPS

Transformation

Method of assessment

Assessed in field from standing water Finding locality in Google Earth Assessed in field

Water extraction of soil, using conductivity metre Sums of measured concentrations of ammonium, nitrite and nitrate, converted towards mol N, converted back into mass N Bray I method Using 1:10 water extraction of soil

Vegetation data is used to obtain a classification of wetland vegetation that groups similar vegetation plots together so that a vegetation typology can be obtained. In this study, the programme PC-Ord was used to obtain such a classification (McCune and Mefford, 2011). A clustering of vegetation plots is achieved by calculating a similarity index for each pair of plots that can be used in subsequent analyses (McCune and Grace, 2002; Legendre and Legendre, 2012). Hierarchical agglomerative cluster analysis uses this distance measure combined with a linkage method that serves as a criterion that helps to decide whether two existing groups of plots should be joined. The linkage method that was applied in the current study is called Ward's linkage method and makes use of an objective function that indicates the amount of information lost in the event when two groups are joined. The clustering algorithm proceeds until all plots are joined in a single cluster. The clustering can be illustrated in the form of a dendrogram that shows how the different groups of plots relate to one another. The level of detail in a dendrogram is determined by the number of clusters that are recognized as separate entities. This number of clusters can be optimized with the help of a second type of analysis, which is Indicator Species Analysis (ISA, Dufrêne and Legendre, 1997). Indicator Species Analysis is used to determine which species are more likely to be found in one cluster compared to the others, so it measures the ‘fidelity’ of a species to a certain cluster. For each species an indicator value (IV) is calculated which is the product of two values: the relative abundance (which represents the proportion of plots in cluster k in which species j occurs) and the relative frequency (which represents the proportion of total occurrences of species j that occur in cluster k). The significance

log

n.a.

Organic

mass %

log

%Carbon

n.a. cm m degrees mS/cm

*

log

PH Inundation Altitude Slope EC

mg/kg

log

Nitrogen

mg/kg mg/kg

log Standardized, then log

Phosphorus Na, Ca, K, Mg

of these Indicator Values is tested by means of a Monte Carlo permutation test, and only those species that have an Indicator Value higher than 20 and a p-value for the permutation test lower than 0.05 should be regarded as real indicator species. The sum of the p-values of all species in the dataset should be as low as possible and in this way ISA can help to establish the number of clusters that should be used in cluster analysis, by repeating the clustering algorithm with a different number of clusters (Dufrêne and Legendre, 1997). The third type of analysis that was carried out on the dataset was canonical ordination, which correlates multivariate data such as species composition in a plot to a set of explanatory variables. The programme that was used for this analysis was Canoco 4.5 (Ter Braak and Šmilauer, 2002) and the type of analysis chosen was Canonical Correspondence Analysis. This technique finds the subsequent axes of maximum variation by means of eigen-analysis of a Chi-square distance matrix of the original dataset and correlates these to the matrix of explanatory variables by means of multiple regression (Ter Braak, 1987; Legendre and Legendre, 2012). The total amount of variation in the dataset as a whole is indicated as Total Inertia and is the sum of all eigenvalues of all axes. The proportion of this that is explained by the environmental variables supplied is also referred to as the Total Variation Explained or TVE and is the sum of all canonical eigenvalues divided by the Total Inertia (Ter Braak, 1987; Legendre and Legendre, 2012). Obviously, the fraction of Total Variation Explained is higher when more environmental variables are supplied. For this reason, a subset of the overall dataset was analysed that includes only those plots that have soil data. Canonical ordination was carried out separately for this dataset.


161

Fig. 1. Map indicating the distribution of subtropical wetland vegetation in the country. The black dots represent historical data, some of which also included soil data, particularly in the Western Cape.

Results

3. Description of communities

The distribution of vegetation plots in the overall dataset of subtropical wetland vegetation is shown in Fig. 1. Subtropical wetlands are mainly found in the coastal regions from the Western Cape up to KwaZulu-Natal and in the northern part of the country. A few aberrant plots in the inland parts of the Eastern Cape, Free State and Northwest Province also should be regarded as Subtropical Wetland Vegetation based on the dominant species. The total dataset of subtropical wetland vegetation consists of 1112 vegetation plots with a total of 1054 species.

The sixteen community groups are listed in Table 4 with regards to their occurrence in the country, their vegetation composition and structure as well as their environmental drivers. A small number of communities occur over a broad range of altitudes, notably the iKhwane wetlands, but most communities are strictly coastal as they are only found at the lowest altitudes, namely the Coastal grassy wetlands (Community group 2) and the Pondo coastal wetlands (Community group 12). A limited number of vegetation types are found exclusively at the higher altitudes, such as the Nylsvley wetland (Tarboton, 1987) which contains some unique vegetation types for the country: these are the Rice grass wetlands (Community group 3) and the Escarpment short sedge wetlands (Community group 8). Altitude influences temperature (which can be compensated by latitude) as well as humidity. The coastal areas are generally more humid than the tropical areas on the Highveld and therefore contain different vegetation.

1. Classification Hierarchical cluster analysis initially led to a classification of 49 clusters which are illustrated in Fig. 2. The left side of this figure represents a summary of the clusters in sixteen distinct groups. Indicator values have been calculated for all species for the sixteen main groups and the indicator species and their indicator values are shown in Table 3.

2. Ordination Soil data from Subtropical Wetland Vegetation is only available for a small fraction of the total dataset, which means that the subset used for ordination analysis is relatively small. The dataset of those plots that contain soil data consists of only 160 plots which cover only 12 of the 16 community groups recognized in Fig. 2 and Table 3, one of which (Fever tree wetlands) is only represented by a single plot. Maputaland is actually under-represented when it comes to plots with soil data, even though there is a lot of historical data available from this area. The results of the CCA ordination are shown in Fig. 3 and this represents an ordination that has a Total Inertia (TI) of 40.818, while the sum of all Canonical Eigenvalues is 8.039. This means that the Total Variation Explained (TVE) is about 19.7%, which is the fraction of the Total Inertia contributed by the canonical eigenvalues.

Discussion The ordination revealed that only 19.7% of the overall variation in the dataset could be explained by the environmental variables supplied. This may reflect that there are environmental variables that have not been included in the study, such as climate or other spatial aspects. The first axis of the ordination was negatively correlated with clay content, with a number of communities associated with a low clay content. These are communities found on the coastal sands, namely the Eleocharis limosa communities, the Coastal grassy communities, the Cladium mariscus communities (also often found in peaty soils) and Cyperus textilis communities. The rest of the communities were associated with high clay content and higher soil electrical conductivity. Some of these communities also have a high soil organic matter content, because clay soils drain slower and have less aerobic decomposition of organic matter (Wattel-Koekkoek et al., 2001). The second axis of the ordination diagram was not directly correlated with a single variable but had a weak positive correlation with inundation. It was also negatively correlated with wetness, which would

162


Fig. 2. Dendrogram showing the main subdivisions in the subtropical wetlands, and how they correspond with the classification reported by Sieben et al. (2014). The number in brackets represents the total number of plots that have been recorded of that type.

E.J.J. Sieben et al. / South African Journal of Botany 104 (2016) 158–166 Table 3 Indicator species analysis for the Subtropical Wetland Vegetation. This table is summarized for the larger clusters defined in Fig. 2.

No. Wetland vegetation type

Indicator species

1 2

Eleocharis limosa wetlands Coastal grassy wetlands

3

Rice grass wetlands

4

Fever tree wetlands

5

Cyperus textilis wetlands

6

Cyperus sexangularis wetlands

7

Ikhwane wetlands

8 9 10

Escarpment short sedge wetlands Cyperus dives wetlands Escarpment peat wetlands

Eleocharis limosa Stenotaphrum secundatum Cyperus sphaerospermus Centella asiatica Lactuca inermis Eleocharis acutangula Oryza longistaminata Acacia xanthophloea Brachiaria deflexa Justicia betonica Euclea divinorum Abutilon angulatum Maytenus heterophyllus Cyperus textilis Zantedeschia aethiopica Cyperus sexangularis Eriochloa meyeriana Ischaemum afrum Setaria incrassata Malvastrum coromandelianum Commelina benghalensis Cyclosorus interruptus Cyperus latifolius Kyllinga erecta

11

Giant reed wetlands

12

Pondo coastal wetlands

13 14

Cladium mariscus wetlands Papyrus wetlands

15

Antelope grass wetlands

16

Mfabeni peat wetlands

Cyperus dives Thelypteris confluens Nidorella auriculata Phragmites mauritianus Ipomoea mauritiana Abildgaardia hygrophila Scleria sobolifer Rhynchospora brownii Smilax anceps Digitaria eriantha Conostomium natalense Trachypogon spicatus Cladium mariscus Cyperus papyrus Cyperus prolifer Pycreus mundii Typha capensis Phragmites australis Persicaria decipiens Ficus verruculosa Echinochloa pyramidalis Sorghum bicolor Persicaria senegalensis Ipomoea cairica Rhynchospora holoschoenoides Rhynchospora corymbosa Eleocharis dulcis Fimbristylis bivalvis Scleria poiformis Schoenoplectus brachyceras Sphagnum truncatum Cynodon dactylon Panicum glandulopaniculatum Xyris natalensis Pycreus nitidus

Indicator value (IV) p-Value 78.9 21.2 19.9 16.3 15.9 16.4 14.3 91.7 74.9 37.5 25 22.8 18.8 80.3 16.1 65.2 44.1 26 21.7 15.2

0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001

52.3 42.1 16.2 18

0.001 0.001 0.001 0.001

59.1 38.7 29.8 57.2 20.2 42.9 39.4 26.1 22.6 19.6 18.9 17.7 90.2 50.5 30.9 29.2 24 20.3 17.1 16.9 85.3 25.4 15.5 15.2 76.8

0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001

69.3 61.1 51.4 35.2 27.7

0.001 0.001 0.001 0.001 0.001

21.6 17.5 16.2

0.001 0.001 0.001

15.7 15.9

0.001 0.001

normally be strongly correlated with inundation. It seems that the wetness gradient was not very strong, possibly because the drier (temporarily wet) part of the gradient is not represented by many plots and they are scattered across the diagram. The Pondo coastal wetlands had the highest nitrogen contents and the Escarpment short grassy wetland vegetation had the lowest nitrogen contents. The high altitude

163

floodplains such as Nylsvley and other floodplains on the Limpopo Highveld were associated with altitude and organic carbon content while the wetlands from the Pondoland were associated with the coast. The wetlands in the subtropical areas of South Africa are similar to the wetlands of the other warm areas of the continent. The largest of these wetlands in Southern Africa are found in Zambia (Kafue Flats, Bangweolu Swamps), Mozambique (Zambezi River delta), and Botswana (Okavango Delta) but only limited studies of the vegetation of these wetlands exist (Ellenbroek, 1987; Ellery et al., 1995, 2001). The largest wetland system in subtropical areas in South Africa is the St. Lucia system with the associated Mkhuze Swamp located in the Maputaland Center of Endemism (Goodman, 1987; Van Wyk and Smith, 2001). Many of the communities that may be rare in South Africa are probably quite common across the borders of the country, for example the Papyrus swamps which are very well documented in the Okavango Delta in Botswana and in East Africa (Ellery et al., 1995; Terer et al., 2015). This may also apply to the wetlands of Community group 16, which, within South Africa, are restricted to Mfabeni Swamp. It is however clear from the current study that, within South Africa, the change in growth form (from small sedges up to 50 cm tall, to the tall sedges that are typical for the tropical swamps in East Africa) takes place along an altitudinal gradient, similar to what has been concluded by Kotze and O'Connor (2000). Communities where large sedges such as Cyperus papyrus, Cyperus dives, Cyperus latifolius and Cladium mariscus are dominant occur within South Africa only in warm coastal areas, but are more common in tropical parts of Africa (Thompson and Hamilton, 1983). Maputaland had the highest diversity of wetland vegetation types for the subtropical regions. This area also has the largest extent of wetland area and is already long known for its conservation value for wetlands (Van Wyk and Smith, 2001; Pretorius, 2011). Unfortunately within the scope of the current project, detailed soil data for most wetlands in Maputaland is not available, as most of the data in the Wetland vegetation database for this area is from historical data that was not collected with the standardized environmental information. There is however a prospect of more information on this area coming available in the near future as Pretorius (2011) has been collecting very detailed data on vegetation patterns and soil variables in recent years. Wetland vegetation data forms an important contribution towards strategic conservation planning (Rivers-Moore and Goodman, 2010; Rivers-Moore et al., 2011) and the current study highlights the large variation found in wetland vegetation in the coastal and warm areas of South Africa. The variation is higher than the range of vegetation types described under ‘(Azf6) Subtropical Freshwater wetlands’ in Mucina and Rutherford (2006) as some of the dominant species described in this study were omitted; namely Cyperus dives, prolifer, C. sphaerospermus, Abildgaardia hygrophila, Rhynchospora corymbosa, R. holoschoenoides, Typha domingensis and Eleocharis limosa. At the same time, some of the species listed in Mucina and Rutherford (2006) are more widespread in South Africa and have been included in the Temperate grassy wetland vegetation (Sieben et al., 2014). It is not clear which of the vegetation types are unique to South Africa, as similar wetlands occur in other parts of the continent, but the assumption is that those wetland vegetation types that are restricted to the KwaZuluNatal and Eastern Cape coastlines, will not be widespread in the rest of Africa. These areas are recognized as belonging to Centres of Endemism (the Albany-Tongaland Region which includes the Pondoland Center of Endemism and the Maputaland Center of Endemism, which extends into Southern Mozambique, Van Wyk and Smith, 2001). For this reason, the Pondo coastal wetland vegetation (Community Group 12) and Mfabeni peat wetland vegetation (Community Group 16) as well as part of the Coastal grassy wetland vegetation (Community Group 2) have the highest importance for biodiversity and conservation planning in South Africa and should be given priority in terms of conservation efforts.

164


Fig. 3. CCA ordination diagram for the first two axes for vegetation data from subtropical wetlands.

The threats to subtropical freshwater wetlands are also highest in coastal KwaZulu-Natal, where continuing coastal development and conversion towards agricultural land are a persistent threat. It has been recognized for a long time that wetlands form valuable habitats that have been declining in South Africa over the last fifty years (Cowan, 1995; Kotze et al., 1995), particularly because of human utilization. Subsistence agriculture in wetlands is an important factor in many of the most densely inhabited areas of South Africa, (the former apartheid homelands of Transkei, KwaZulu, Lebowa, Gazankulu and Venda) where most of these wetlands are found. Since the wetlands provide the most fertile environment in areas such as Maputaland, many wetlands are ploughed up for the growing of crops, particularly Taro (Colocasia esculenta), which prefers a high soil moisture content (Saunders et al., 2012). Problems arise when people drain the wetlands for the cultivation of other crops such as banana, sweet potato, cabbage or the planting of woodlots. The wise use that allows for sustainable management of the wetlands in these areas represents one of the biggest challenges in wetland conservation in South Africa. Those wetlands that are found in a productive environment that are subjected to shifting cultivation are more resilient in the long term as the natural vegetation quickly grows back after cultivation has ended. It is however mostly those wetlands in the unproductive habitats of coastal sands (Pondo coastal wetland vegetation and Mfabeni peat wetland vegetation as well as some of the coastal grassy wetland vegetation types) that are likely to be severely damaged after cultivation takes place. The succession of these wetlands after disturbance and their resilience is an important and interesting area for further research. The productivity of the wetlands does not only benefit the production of crops but also the growth of tall indigenous vegetation, many of which can be utilized as fibres for construction and traditional crafts. For this reason, many of the wetlands in subtropical areas should also be regarded as having a high cultural significance (Kotze and Traynor, 2011). Transitional stages between subtropical wetland vegetation and Swamp Forest have been found in a number of locations in northern KwaZulu-Natal and it has been suggested that where the topography protects the wetlands against fire, many of the vegetation types would develop into Swamp Forest (Grobler, 2009; Wessels, 1997). This is already happening in some places around the Mfolozi River, where a new channel has been dug that isolates an existing wetland from the

surrounding savannas (Wessels, 1997). The successional link between subtropical wetland vegetation and swamp forest is particularly relevant since Swamp Forest is regarded as a threatened vegetation type (Mucina and Rutherford, 2006; Wessels, 1997). Many Swamp Forests have also been targeted for community gardening and the growing of crops (Grobler, 2009). In some areas where Swamp Forest is expanding due to exclusion of fire, a conservation concern is the loss of low sedge vegetation that is being replaced, as it is becoming clear that many of these vegetation types are just as rare as the Swamp Forest itself (Luvuno, 2013). Many of the wetlands in arid savannah ecosystems that are rich in large grazing animals develop grass lawn vegetation in response to excessive grazing and trampling (McNaughton, 1979, 1984). It is for this reason that many wetlands in subtropical areas also belong to the group of grass lawn wetlands (Sieben et al., in preparation). The successional series between lawn grass wetlands and subtropical wetlands is also particularly relevant as many cattle are wandering freely in the communal areas of Northern KwaZulu-Natal and the Escarpment areas of Mpumalanga and Limpopo. There are certainly management challenges around subtropical wetland vegetation in South Africa, mainly associated with the small scale of many of these wetlands and the ecosystem services that they provide for a large rural population. Excessive utilization of these wetlands often leads to degradation which can in turn have a detrimental effect on rural livelihoods (Schuyt, 2005; Jogo and Hassan, 2010). Therefore, knowledge of the indicators for a healthy ecosystem is a useful tool in the long-term management of wetlands in these areas.

Acknowledgements This work was made possible by the generous funding of the Water Research Commission, Project K5/1980. We would like to thank those people who have been prepared to make historical data available for analysis in the national wetlands database, notably Fred Ellery, Greg Mullins, Mbali Goge and Retief Grobler. We also would like to thank our friend and colleague Thilivhali Nyambeni, who sadly passed away during the drafting of this manuscript and with whom we spent many hours in the field. His contribution to the dataset is significant as he surveyed the whole of the Limpopo province for wetland vegetation.


165

Table 4 Description of the sixteen community groups. Wetland vegetation No. type

No. comm. Dominant species

Vegetation structure

Environmental conditions

1

Eleocharis limosa wetlands

1

Eleocharis limosa

Dense, medium tall sedges (50–80 cm)

Grey clay soils that are seasonally to permanently inundated

2

Coastal grassy wetlands

8

Dense, short lawn grasses and sedges (20–50 cm)

Pans and valley bottoms, temporary to seasonally wet, mostly on soils rich in sand.

3

Rice grass wetlands

4

Dense, short grasses (20–50 cm tall)

Permanently wet soils, often on peat

4

Fever tree wetlands

1

Stenotaphrum secundatum, Cyperus sphaerospermus, Pycreus polystachyos, Centella asiatica Leersia hexandra, Oryza longistaminata Acacia xanthophloea, Brachiaria deflexa

5

1 Cyperus textilis wetlands 3 Cyperus sexangularis wetlands

Cyperus textilis

Cyperus sexangularis, Eriochloa meyeriana

Short lawn grasses and sedges combined with tall trees. Dense, tall sedge community (100–150 cm) Dense, tall sedge community (100–150 cm)

7

Ikhwane wetlands

3

Cyperus latifolius, Cyclosorus interruptus

Dense, tall sedgeland (100–150 cm)

Edges of pans and floodplains, temporarily wet. Clay soils on the edge of pans and valley bottoms, mostly seasonally wet Drier temporarily wet edges of pan or valley bottom wetlands, mostly on black clay soils Clay soils that are permanently wet, mostly valley bottom wetlands

8

Escarpment short sedge wetlands

4

Kyllinga erecta, Pycreus mundii, Pycreus nitidus

Dense, short to medium tall sedgeland (20–80 cm)

9

Cyperus dives wetlands Escarpment peat wetlands

3

Cyperus dives

2

Thelypteris confluens, Trichopteryx dregeana

Dense tall sedgeland Mostly on temporarily to (100–150 cm) seasonally wet clay substrates Permanently wet, often on Dense short mixed peat substrates grassland and herbland (20–50 cm) Open to dense very Often in floodplains, in heavy clay soils that can tall reedland dry out regularly (150–250 cm)

6

10

Mostly in seasonally to permanently wet clay soils

11

Giant reed wetlands

2

Phragmites mauritianus

12

Pondo coastal wetlands

5

Abildgaardia hygrophila, Scleria sobolifer, Digitaria eriantha

Dense short sedgeland (20–80 cm)

Permanently wet substrates, often on peat, mostly in very nutrient-poor substrates

13

Cladium mariscus wetlands

2

Cladium mariscus

Dense very tall sedgeland (150–250 cm)

Often coastal settings, on loamy soils, permanently wet

14

Papyrus wetlands

2

Cyperus papyrus, Cyperus prolifer

Dense very tall sedgeland (150–250 cm)

Permanently wet soils, mostly on peat.

15

Antelope grass wetlands Mfabeni peat wetlands

2

Echinochloa pyramidalis

7

Rhynchospora holoschoenoides, Rhynchospora corymbosa, Scleria poiformis, Panicum glandulopaniculatum

Dense very tall grassland (150–150 cm) Dense short to medium short sedgeland (20–120 cm)

Floodplains of major rivers, mostly on temporarily to seasonally wet loam soils. Peat substrate, permanently wet

16

References Boutin, C., Keddy, P.A., 1993. A functional classification of wetland plants. Journal of Vegetation Science 4, 591–600.

Species diversity

Distribution

Comments

Mostly on the coastal plains of the Cape, but also along the Eastern escarpment Mostly coastal wetlands, but also some communities in Limpopo.

Median Replaced with the similar Eleocharis dregeana wetlands in 4 (2–14) the more temperate inland areas Median 9 (1–28)

Heterogeneous group of Maputaland, Mpumalanga lowveld and wetland communities Nylsvley in Limpopo Maputaland and Limpopo

Western Cape coast

Mostly in Limpopo, but also found in coastal KwaZulu-Natal Mostly in coastal KwaZulu-Natal and Eastern Limpopo, but extending down towards the Cape (Plettenberg Bay) Inland areas and along the escarpment of Eastern Cape, KwaZulu-Natal, Mpumalanga and Limpopo Coastal KwaZulu-Natal and eastern parts of Limpopo Mpumalanga escarpment

Limpopo, the Mpumalanga Lowveld and Northern KwaZulu-Natal Mostly in Pondoland, but some extending into similarly poor substrates on the coastal sands of Northern KZN and the Mpumalanga escarpment Mostly coastal, especially in the Eastern Cape, but also in Northwest province In South Africa only in Northern KwaZulu-Natal, but widespread throughout tropical Africa. In South Africa restricted to Northern KwaZulu-Natal Northern KwaZulu-Natal

Often a monoculture as the species grows very densely on its rhizome. Similar in appearance to previous community

Median 6 (1–20) Median 7 (3–13) Median 3 (1–10) Median 6 (1–12)

This wetland type is economically Median 9 (2–33) important as it contains the heavily utilized species Ikhwane (Cyperus latifolius) Median 9 (2–23)

Occurs in similar areas as those of Community group 7, but in slightly drier areas.

Median 7 (2–12) Median 11 (6–18)

Similar to the community of common reed (Phragmites australis), but better coping with drought stress. In Pondoland, these are unique communities that are often just as rich in species as the surrounding uplands.

Median 7 (1–17)

Very tall reedlands which are quite impenetrable.

Median 5 (1–15)

Known to create floating islands (‘Sudd’) in the Okavango Delta that block river channels

Median 8 (1–15)

Median 12 (2–27)

Median 7 (1–16) In South Africa, these communi- Median 5 (2–17) ties are mostly restricted to a single wetland: Mfabeni Swamp. It is unknown how widespread these communities are in neighbouring Mozambique.

Brown, L.R., Du Preez, P.J., Bezuidenhout, H., Bredenkamp, G.J., Mostert, T.H.C., Collins, N.B., 2013. Guidelines for phytosociological classification and description of vegetation in southern Africa. Koedoe 55. http://dx.doi.org/10.4102/koedoe. v55i1.1103.

166


Coetzee, J.P., Bredenkamp, G.J., Van Rooyen, N., 1994. Phytosociology of the wetlands of the Ba and Ib land types in the Pretoria-Witbank-Heidelberg area of the Transvaal. South African Journal of Botany 60, 61–67. Collins, N.B., 2011. Phytosociology and ecology of selected depression (pan) and valleybottom wetlands of the Free State Province PhD Thesis University of the Free State, Bloemfontein. Corry, F.T.H., 2011. The development of a macrophyte based index of environmental condition of inland wetlands from the coastal lowlands of the Western Cape of South Africa. Water Research Commssion Report K5/1584. Cowan, G.I., 1995. Wetland regions of South Africa. In: Cowan, G.I. (Ed.), Wetlands of South Africa. Department of Environmental Affairs and Tourism, Pretoria, pp. 21–32. Cowden, C., Ellery, W.N., Kotze, D.C., Sieben, E.J.J., 2014. The impact or restoration on two wetlands in KwaZulu-Natal, South Africa. African Journal of Aquatic Sciences 39, 237–247. Cronk, J.K., Fennessey, M.S., 2001. Wetland Plants: Biology and Ecology. Lewis Publishers, Boca Raton 462 pp. Driver, A., Nel, J.L., Snaddon, K., Murray, K., Roux, D.J., Hill, L., Swartz, E.R., Manuel, J., Funke, N., 2011. Implementation Manual for Freshwater Ecosystem Priority Areas. WRC Report No. 1801/1/11. Dufrêne, M., Legendre, P., 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Monographs 67, 345–366. Ellenbroek, G.A., 1987. Ecology and productivity of an African wetland system: the Kafue Flats, Zambia. Geobotany Series. Junk, The Hague 267 pp. Ellery, W.N., Ellery, K., Rogers, K.H., McCarthy, T.S., 1995. The role of Cyperus papyrus L. in channel blockage and abandonment in the Northeastern Okavango Delta, Botswana. African Journal of Ecology 33, 25–49. Ellery, W.N., McCarthy, T.S., Dangerfield, J.M., 2001. Floristic diversity in the Okavango Delta, Botswana as an endogenous product of biological activity. In: Gopal, B., Junk, W.J., Davis, J.A. (Eds.), Biodiversity in Wetlands: Assessment, Function and Conservation volume 1. Backhuys, Leiden, pp. 195–226. Furness, H.D., 1981. The Plant Ecology of Seasonally Flooded Areas of the Pongolo River Floodplain with Particular Reference to Cynodon dactylon (L) Pers PhD Thesis University of KwaZulu-Natal. Goge, M.C., 2002. Classification of Wetlands of the Greater St. Lucia Wetland Park: A Hydrogeomorphic Approach MSc Thesis University of KwaZulu-Natal. Goodman, P.S., 1987. Mkhuze floodplain and swamps—wetland management study. African Wildlife 41, 1–5. Grobler, L.E.R., 2009. A Phytosociological Study of Peat Swamp Forests in the Kosi Bay Lake System, Maputaland, South Africa MSc Thesis University of Pretoria. Hennekens, S.M., Schaminée, J.H.J., 2001. Turboveg, a comprehensive database management system for vegetation data. Journal of Vegetation Science 12, 589–591. Jogo, W., Hassan, R., 2010. Balancing the use of wetlands for economic well-being and ecological security: the case of the Limpopo wetland in Southern Africa. Ecological Economics 69, 1569–1579. Kareko, J., 2002. The Vegetation of Middelvlei, Stellenbosch Masters Thesis University of Stellenbosch. Keddy, P.A., 2004. Wetland ecology. Principles and conservation. Cambridge Studies in Ecology Cambridge. 614 pp. Kotze, D.C., O'Connor, T.G., 2000. Vegetation variation within and among palustrine wetlands along an altitudinal gradient in KwaZulu-Natal, South Africa. Plant Ecology 146, 77–96. Kotze, D.C., Traynor, C.H., 2011. Wetland plant species used for craft production in KwaZulu-Natal, South Africa: ethnobotanical knowledge and environmental sustainability. Economic Botany 65, 271–282. Kotze, D.C., Breen, C.M., Quinn, N., 1995. Wetland losses in South Africa. In: Cowan, G.I. (Ed.), Wetlands of South Africa. Department of Environmental Affairs and Tourism, Pretoria. Kotze, D.C., Hughes, J.C., Klug, J.R., Breen, C.M., 1996. Improved criteria for classifying hydric soils in South Africa. South African Journal of Plant and Soil 13, 67–73. Kotze, D.C., Marneweck, G.C., Batchelor, A.L., Lindley, D.S., Collins, N.B., 2008. Wetecoservices. A technique for rapidly assessing ecosystem services supplied by wetlands. WRC Report TT 339/08. Legendre, P., Legendre, L., 2012. Numerical Ecology. third ed. Elsevier, Amsterdam 990 pp. Luvuno, L., 2013. Burning wetlands: the influence of fire on wetland vegetation, structure and composition MSc Thesis University of KwaZulu-Natal. McCune, B., Grace, J.B., 2002. Analysis of Ecological Communities. MjM Software Design, Oregon. McCune, B., Mefford, M.J., 2011. PC-ORD. Multivariate Analysis of Ecological Data, Version 6. MjM Software, Gleneden Beach, Oregon. McNaughton, S.J., 1979. Grazing as an optimization process: grass-ungulate relationships in the Serengeti. American Naturalist 113, 691–703. McNaughton, S.J., 1984. Grazing lawns: animals in herds, plant form and coevolution. American Naturalist 124, 863–886. Mitsch, W.J., Gosselink, J.G., 2000. Wetlands. Third edition. Van Nostrand Rienhold, New York, USA.

Mucina, L., Rutherford, M.C., 2006. The vegetation of South Africa, Lesotho and Swaziland. Strelitzia 19 Pretoria. Neal, M., 2001. The Vegetation Ecology of the Lower Mkuze River Floodplain, Northern KwaZulu-Natal: A Landscape Ecology Perspective Masters thesis University of Natal, Durban. Nel, J.L., Murray, K.M., Maherry, A.M., Petersen, C.P., Roux, D.J., Driver, A., Hill, L., Van Deventer, H., Swartz, E.R., Smith-Adao, L.B., Mbona, N., Downsborough, L., Nienaber, S., 2011. Technical report for the National Freshwater Ecosystem Priority Areas Project. WRC Report No. 1801/2/11. Ollis, D., Snaddon, K., Job, N., Mbona, N., 2013. Classification system for wetlands and other aquatic ecosystems in South Africa. User Manual: Inland Systems. SANBI Biodiversity series 22. Perkins, L., Bredenkamp, G.J., Granger, J.E., 2000. Wetland vegetation of Southern KwaZulu-Natal, South Africa. Bothalia 30, 175–185. Pretorius, M.L., 2011. A vegetation classification and description of five wetland systems and their respective zones on the Maputaland Coastal Plain. MSc Thesis. University of South Africa, Pretoria. Rivers-Moore, N.A., Goodman, P.S., 2010. River and wetland classifications for freshwater conservation planning in KwaZulu-Natal, South Africa. African Journal of Aquatic Sciences 35, 61–72. Rivers-Moore, N.A., Goodman, P.S., Nel, J.L., 2011. Scale-based freshwater conservation planning: towards protecting freshwater biodiversity in KwaZulu-Natal, South Africa. Freshwater Biology 56, 125–141. Rutherford, M.C., Westfall, R.H., 1986. Biomes of Southern Africa—an objective categorization. Memoirs of the Botanical Survey of South Africa 54, 1–98. Saunders, M.J., Kansiime, F., Jones, M.B., 2012. Agricultural encroachment: implications for carbon sequestration in tropical African wetlands. Global Change Biology 18, 1312–1321. Schlesinger, W.H., Bernhardt, E.S., 2013. Biogeochemistry. An Analysis of Global Change. Elsevier, Academic Press, Amsterdam. Schoultz, A., 2000. Plant Community Classification and Environmental Gradient Correlates along the Eastern Portion of the Mkuze Swamps MSc Thesis University of KwaZulu-Natal. Schuyt, K.D., 2005. Economic consequences of wetland degradation for local populations in Africa. Ecological Economics 53, 177–190. Shipley, B., Keddy, P.A., Moore, D.R.J., Lemky, K., 1989. Regeneration and establishment strategies of emergent macrophytes. Journal of Ecology 77, 1093–1110. Sieben, E.J.J., W.N. Ellery, S. Garden, S., Grenfell, M., 2006. Unpublished vegetation data for the wetlands of Zulti South, Richard's Bay. Sieben, E.J.J., Kotze, D.C., Morris, C.D., 2010. The vegetation of the wetlands in the MalotiDrakensberg, South Africa. Bothalia 40, 117–134. Sieben, E.J.J., Mtshali, H., Janks, M., 2014. National Wetland Vegetation Database: Classification and analysis of wetland vegetation types for conservation planning and monitoring. Water Research Commission Report K5/1980. Tainton, N., 1999. Veld Management in South Africa. University of Natal Press, Pietermaritzburg. Tarboton, W.R., 1987. The Nyl floodplain: its significance, phenology and conservation status. In: Walmsley, R.D., Botton, M.L. (Eds.), Ecology and Conservation of Wetlands in South Africa. CSIR, Pretoria, pp. 101–114. Taylor, C.A., 2000. The Relationship between Plant Community Distribution and Environmental Gradients of the Totweni Pan. University of Natal, Durban, Honours Project. Ter Braak, C.J.F., 1987. The analysis of vegetation-environment relationships by canonical correspondence analysis. Vegetatio 69, 69–77. Ter Braak, C.J.F., Šmilauer, P., 2002. Canoco 4.5 Reference manual and CanoDraw for Windows User's Guide. Software for Canonical Community Ordination. Biometris, Wageningen. Terer, T., Muasya, A.M., Triest, L., 2015. Strong isolation by distance revealed among Cyperus papyrus populations in the Rift Valley lakes, Lake Victoria, and isolated wetlands of Kenya. Aquatic Botany 121, 57–66. Thompson, K., Hamilton, A.C., 1983. Peatlands and swamps of the African Continent. In: Gore, A.J.P. (Ed.), Ecosystems of the World 4B. Mires: Swamp, Bog, Fen and Moor. Elsevier, Amsterdam, pp. 331–373. Van Wyk, A.E., Smith, G.F., 2001. Regions of Floristic Endemism in South Africa: A Review with Emphasis on Succulents. Umdaus Press, Pretoria. Venter, C.E., 2002. The Vegetation Ecology of Mfabeni Peat Swamp. University of Pretoria, St. Lucia MSc Thesis. Wattel-Koekkoek, E.J.W., Van Genuchten, P.P.L., Buurman, P., Van Lagen, B., 2001. Amount and composition of clay-associated soil organic matter in a range of kaolinitic and smectitic soils. Geoderma 99, 27–49. Wessels, N., 1997. Aspects of the ecology and conservation of swamp forests in South Africa. MSc Thesis in Forestry Faculty, Saasveld, Port Elizabeth technikon. Westhoff, V., Van der Maarel, E., 1978. The Braun-Blanquet approach. In: Whittaker, R.H. (Ed.), Classification of plant communities, pp. 287–399 Dr. W. Junk, The Hague.