Invasive alien woody plants of the eastern Cape

The frequency and abundance of invasive alien woody plants were recorded along roadsides and at watercourse crossings in 69.9% (151/216) of the quarter degree squares in the study area. The survey yielded 101 species of which the most prominent (in order of prominence) in roadside and veld habitats were: Opuntia ficus-indica, Acacia meamsii and A. cyclops. The most prominent species (in order of prominence) in streambank habitats were: A. meamsii, Populus x canescens, Salix babylonica and S. fragilis (fide R.D. Meikle). The greatest intensity of invasion was recorded in the wetter eastern parts and particularly in the vicinity of Port Elizabeth. Uitenhage, East London, Grahamstown, Hogsback and Stutterheim. There was relatively little invasion in the central and western dry interior except along watercourses.


Survey history and objectives
This study of the eastern Cape is the fifth of eight regional surveys which together are designed to reflect invasion by woody alien plants in the Republic of South Africa as a whole. Surveys have been completed for the Transvaal (Henderson & Musil 1984), Natal (Henderson 1989), Orange Free State (Henderson 1991a) and northern Cape (Henderson 1991b). This survey of the eastern Cape was undertaken in March, October andNovember 1988 andMarch 1990.
The objectives of the survey are: to produce a checklist of the major invasive alien woody plants of streambank, roadside and veld habitats in the study area; to determine the pattern of alien woody invasion as a whole and for in dividual species; to attempt to relate distribution to environmental factors and to determine which are the most prominent and potentially important invaders.

The study area
The study area lies between latitudes 30° and 34°S and 23° and 29°E (Figure 1). The altitude rises in successive terraces from sea level on the Indian Ocean in the south and southeast to 3 000 m in the Drakensberg in the north east. Four major physical divisions can be delimited (Nicol 1988). These are the coastal subregion stretching inland to the 300 m contour; the southern coastal mountains up to 1 500 m high lying west, north and northeast of Port Elizabeth; the midland region which is hilly to moun tainous country and includes the Winterberg with a max imum height of 2 369 m; and the northern mountain region which extends from the Sneeuberg in the west to the Stormberg and Drakensberg in the east. Seven major river systems arise in, and drain, the study area.
Rainfall ranges from 150 mm per annum in the extreme western interior to 1 700 mm in the Amatole Mountains (Dent et al. 1989). Most of the western and central regions receive less than 500 mm per annum (Kopke 1988). The seasonal distribution of rainfall ranges from a winter maximum on the coast between Port Elizabeth and Pbrt Alfred through to a summer maximum in the northern interior (Kopke 1988).
Temperatures vary greatly from the coast inland. The coastal zone is mild in both winter and summer (Kopke 1988). The climate becomes progressively more temperate towards the arid west and with increasing altitude in the north. The interior above the Winterberg escarpment is characterized by hot summers, cold winters and wide spread frost (Kopke 1988). Snow has been recorded occasionally for a few localities at low altitudes (e.g. Grahamstown) and is regular in mountainous parts (Gibbs Russell & Robinson 1981).
Four major vegetation units or biomes and 21 vegeta tion categories have been described in the eastern Cape by Lubke et al. (1986). For the purposes of this survey and in keeping with previous surveys, the vegetation of the study area has been subdivided according to the biomes of southern Africa defined by Rutherford & Westfall (1986) and Acocks's Veld types of South Africa (1988). The Grass land, Savanna, Fynbos and Nama-Karoo Biomes converge in the eastern Cape ( Figure 2). Twenty-six Acocks Veld Types occur in the study area and have been grouped into seven broad veld type categories for the purposes of this survey (Table 1 and Figure 3).
Temperate grassland occupies the highest and coldest parts of the study area at elevations of 1 500 m to 3 000 m. Rainfall ranges from 300 mm in the west to 1 000 mm in the extreme northeast. Moist subtropical grassland occurs on the cool and wet eastern and southeastern slopes of the Drakensberg at elevations from 600 m to 2 000 m. Rainfall ranges from 500 mm to 1 700 mm. Pockets of Afromontane forest occur in favourable localities.
Coastal 'forest' occupies the mild coastal belt with an annual rainfall ranging from 600 mm in the south to 1 000 mm in the north. Vegetation types occurring in this zone are forest, dune thicket, Acacia savanna, grassland and littoral strand vegetation (Lubke etal. 1986). Subtropical thicket and savanna occurs from sea level to about 1 500 m. Rainfall ranges from 200 mm in the hot and dry river valleys to 900 mm on the foothills of the Winterberg.
Fynbos shrublands, hereafter referred to broadly as mountain fynbos, occur along the tops and slopes of the southern coastal mountains at an altitude ranging from 300 m to 1 500 m. Small oudiers are situated within the Savan na Biome along the Suurberg and on the Grahamstown hills. Rainfall ranges from 500 mm to 900 mm per annum. False karoo, at an altitude of between 1 000 m and 1 500 m, occupies areas formerly covered by grassland. Annual rainfall ranges from 200 mm to 500 mm. Karoo or dwarf shrubland occupies the very arid and western interior at an altitude of between 500 m and 1 000 m with an annual rainfall of between 150 mm and 400 mm.

Sampling method
The method used in this survey was basically the same as that used in previous surveys. The changes to the abundance scale for streambank habitats adopted by Henderson (1991b) have also been followed here (see next subheading). The presence and abundance of all alien trees, large shrubs and conspicuous climbers which appeared to be spreading spontaneously (naturalized) were recorded for each veld type category, habitat type (road sides and adjoining veld, and streambanks) and quarter degree/fifteen minute square traversed by road. Twenty quarter degree squares were selected for more intensive surveying ( Figure 2). They may be used at a later date for a quick resurvey of the study area to assess any changes that may have taken place.
Recordings of roadside and veld invaders were made from a moving vehicle along road transects of between five and ten kilometres in length. The average transect length was 7.3 km for the general survey area and 5.0 km for intensive sites. Recordings of streambank invaders were made at virtually all watercourse crossings on the road transects. Details of the roads traversed are lodged in the P.P.R.I., Pretoria. As on previous occasions the survey was undertaken in a minibus, with one driver and one recorder (the author). The average speed was 60 km/h but ranged from about 20 km/h in densely vegetated areas to 100 km/h in sparsely vegetated areas.

Abundance ratings
The abundance ratings for roadside and veld habitats and streambank habitats are given in Table 2.

Sampling level achieved
The sampling level achieved was 69.9% (151 out of the total 216 quarter degree squares) at an average of 29.9 km travelled per square. An average of 18.5 km of road transects were sampled per quarter degree square for abundance estimates of roadside and veld invaders. The mean surface area of each of the quarter degree squares, in which 20 intensive sampling sites are situated, is 646 km2 (23.39 x 27.62 km).
The veld type coverage in terms of quarter degree squares and road transects sampled, kilometres travelled and watercourse recordings made, is given in Table 3. Statistics for streambank, roadside and veld habitats are given in Tables 4 & 5.

Frequency
The percentage frequency of occurrence of a species x in a given category (veld type, biome or study area) y was calculated as follows: frequency = no. of watercourse recordings/road transects in category y having species x total no. of watercourse recordings/road transects in category y x 100

Prominence value
The prominence value is a combined measure of a species' frequency and abundance relative to that of all other species, within a given vegetation category (veld type, biome or study area).
In streambank habitats the prominence value for a species x in category y was calculated as follows: prominence value = total weighted abundance of species x in category y sum of the weighted abundances of all species in category y + frequency of species x in category y sum frequency of all species in category y x 100 x 100 prominence value = total abundance* of species x in category y sum of the abundances* of all species in category y + frequency of species x in category y sum frequencies of all species in category y x 100 x 100 The highest prominence values in a given category which add up to approximately 160 points out of a total of 200 are printed in bold in Tables 6, 7, 8 and 9. The cut-off point of 160 points is arbitrary but represents 80% of the summed prominence values. Tables 8 and 9) The mean species abundance rating** of a species x in a given category (veld type, biome or study area) y was calculated as follows: mean no. of individuals or groups per 10 km total no. of individuals or groups of species x in category y total distance along which species x was rated in category y

x 10
The abundance ratings were weighted according to the minimum percentage cover in each scale rating (see Table  2). Thus ratings 7, 6, 5 and 4 had weighted values of 75, 50, 25 and 5 respectively. Ratings 1, 2 and 3 each had weighted values of 1.
In roadside and veld habitats the prominence value for a species x in category y was calculated as follows: each abundance rating was expressed in numbers of individuals or groups recorded per transect (see Table 2). To be both conservative and consistent the minimum number was used in each instance, e.g. an abundance rating of 5 over ten kilometres = 50 and an abundance rating of 5 over five kilometres = 25. ' mean no. of individuals or groups per 10 km converted to rating (see Table 2). Table 5)

Mean abundance of invaders per km in roadside and veld habitats (see
The mean abundance of invaders per kilometre in a given category (veld type, biome or study area) y/quarter degree square z was calculated as follows: total abundance* of all species in category y/quarter degree square z mean abundance -[ota] kilometres rated for abundance estimates in category y/quarter degree square z

RESULTS
The survey yielded 101 naturalized alien species. These species are listed in the Appendix together with a further 29 species which were obtained from various literature and other sources. The distributions of 30 of the most prominent species are given in Figures 7 and 8.

The whole study area
Six hundred and thirty-eight watercourse crossings were sampled in which 72 species were recorded, with up to nine species in one sample. Invaders were present at 61.0% of all crossings and 9.1% of all crossings were heavily in vaded (Table 4).

Analysis according to veld type
Invasion was intense in both mountain fynbos and moist subtropical grassland where the highest percentages of river crossings were recorded as invaded and heavily invaded. The greatest number of species was recorded in subtropical thicket and savanna but few crossings were heavily invaded in this veld type category. Overall the Fynbos Biome was the most heavily invaded in terms of percentage crossings invaded and percentage crossings   heavily invaded. The Grassland Biome was the next most heavily invaded followed by the Savanna Biome and lastly the Nama-Karoo Biome (Table 4).

Analysis according to species
Frequency Salix babylonica was the most frequently recorded invader in the study area (19.6%). Only this species and Populus x canescens (11.8%) were recorded at 10% or more crossings in the whole study area (Table 7).
Other species which were recorded at 10% or more crossings in a veld type category were: Acacia cyclops and Eucalyptus spp. in mountain fynbos; Airiplex cf. nummularia in karoo; Acacia dealbata, A. meamsii, Prunus persica and Salix caprea in moist subtropical grassland;

Prominence
The most prominent invader in the whole study area was Acacia meamsii with a prominence value of 32.2 out of a combined total for all species of 200 (Table 7). The next most prominent invaders were Populus x canescens (28.7) and Salix babylonica (28.2).
In the Fynbos Biome Acacia mearnsii was by far the most prominent invader followed by Populus X canescens and A. saligna. In the Nama-Karoo Biome Airiplex cf. nummularia was the most prominent invader in the karoo veld type category. Salix babylonica was the most promi nent invader in false karoo and the whole of the Nama-Karoo Biome.
In the Grassland Biome Salix babylonica, Populus x canescens and S. fragilis were the most prominent in vaders. The same species were also the most prominent invaders in temperate grassland. Acacia meamsii, S. babylonica and A. dealbata were the most prominent invaders in moist subtropical grassland.
In the Savanna Biome Sesbania punicea, Arundo donax, Ricinus communis and Acacia mearnsii were the most prominent invaders. A. cyclops was most prominent in coastal 'forest' and Arundo donax was most prominent in subtropical thicket and savanna.

The whole study area
One hundred and fifty one quarter degree squares and 384 road transects were sampled in which 94 species were recorded. Up to 25 species were recorded per quarter degree square. Naturalized species were recorded in 98.2% of all transects sampled and 28.1% of all transects were heavily invaded (Table 5).

Analysis according to veld type
Invasion was most intense in mountain fynbos where the highest percentage of transects was heavily invaded and the mean abundance of invaders per km reached a maxi mum ( Table 5). The next most heavily invaded categories were coastal 'forest', subtropical thicket and savanna, and moist subtropical grassland. The greatest number of spe cies was recorded in coastal 'forest'.

Prominence
Opuntia ficus-indica scored the highest prominence value of 58.4 in the study area. The next most prominent species were Acacia meamsii (20.8) and A. cyclops (15.2) ( Table 9).
In the Fynbos Biome, Acacia meamsii, A. saligna, A. cyclops and Pinus pinaster were the most prominent species. In the Nama-Karoo Biome, Opuntia ficus-indica was the most prominent species followed by O. cf. robusta cultivars and Agave americana.
In the Grassland Biome. Rosa eglanteria, Acacia meamsii and Opuntia ficus-indica were the most prominent inva ders. In the Savanna Biome, O. ficus-indica was by far the most prominent invader followed by A. cyclops and A. meamsii. Acacia dealbata and Rubus affinis deserve mention as the second and third most prominent invaders after A. meamsii in moist subtropical grassland. Psidium guajava was ranked fourth in coastal 'forest' after A. cyclops, A. meamsii and Opuntia ficus-indica. A. longifolia and Hakea sericea were abundant in places within mountain fynbos.

Patterns of invasion
Alien plant invasion was recorded in streambank, road side and veld habitats throughout the eastern Cape (Figures A comparison of Figures 4 and 5 shows that similar patterns of invasion were recorded in streambank. roadside and veld habitats, except that in the western dry mountain areas there was more severe invasion of the streambank habitat than of roadside and veld habitats. This pattern of streambank invasion in the dry mountain areas was almost entirely due to Populus x canescens ( Figure 8C) and Salix babylonica ( Figure 8K).

Prominent and potentially important species
Several Opuntia species have been, or still are, trouble some invaders in the eastern Cape. O. wilgaris was a major weed at the end of the nineteenth century but today is of minor importance following a very successful biological control programme (Zimmermann et al. 1986). Species infesting large areas at present are O. ficus-indica and    O. aurantiaca (Zimmermann et al. 1986). The latter species, known as jointed cactus, is an inconspicuous lowgrowing species and was excluded from this survey because it was easily overlooked.
Opuntia ficus-indica ( Figure 7N) has been naturalized in the eastern Cape for more than 200 years. According to MacDonald (1891) it was first introduced to this region in 1750. Although it was found growing wild between 1772 and 1775 it seems that until at least 1834 it remained largely within the confines of cultivation. By 1859 it had infested a few farms. Thereafter it spread rapidly and by 1891 it had infested 282 000 ha of land in the districts of Graaff-Reinet, Aberdeen. Jansenville. Somerset East and Willowmore. Localized infestations were found in many other districts. By 1932, prior to a biological control campaign, it occurred on 800 000 ha of land in the Cape Province: 400 000 ha in the eastern Cape and Karoo were densely infested .
Cochineal (Dactylopius opuntiae) aided by felling, caused the collapse of 80% of the 400 000 ha of dense infestations . The moth Cactoblastis cactorum was effective in killing a substantial proportion of the in r*> more isolated plants (Zimmermann et al. 1986). The present distribution of Opuntia ficus-indica is mainly a reflection of the effects of climate on the insect herbivores, particularly Dactylopius opuntiae, and not a direct influence of climate on the plant itself (Zimmermann et al. 1986). The insects are most effective under hot and dry conditions and least effective under cool and moist conditions (Zimmermann et al. 1986). Acacia meamsii ( Figure 7D) was the next most promi nent invader after Opuntia ficus-indica in roadside and veld habitats and the most prominent species in streambank habitats. It was most abundant in the cool and moist regions which support mountain fynbos and moist sub tropical grassland. It was frequently recorded in the warmer coastal lowlands but its average abundance was less than in the previous categories.
Whereas Opuntia ficus-indica is being kept in check by its natural insect herbivores, Acacia meamsii has tremen dous potential to spread. This is largely due to its ability to produce large quantities of long-lived seeds and the absence of natural seed predators. Seed can remain viable for more than 50 years and over 20 000 seeds per square metre can accumulate under an old tree . Seed is very efficiently dispersed by water along water courses, but judging from the dense stands which develop along roadsides, it can also be dispersed in soil by roadbuilding activities and possibly vehicle tyres. I predict that Acacia meamsii will continue to expand its range and that all the cool and moist mountain regions are particularly susceptible to invasion, as well as all watercourses within the Fynbos, Grassland and Savanna Biomes.
In this survey Acacia cyclops ( Figure 7A) was found to be restricted to the coastal lowlands and mountains. It was 24 26 28 heavily invasive in parts of coastal 'forest', subtropical thicket and savanna, and mountain fynbos. It was parti cularly abundant in coastal dune vegetation around Port Elizabeth where it appeared to be the commonest woody species. Its presence in this area dates back to at least the 1890's when it, A. saligna, A. pycnantha and Pinus hale pensis were used in a sand dune reclamation scheme . Taylor & Morris (1981) are of the opinion that A. cyclops threatens to destroy the structure of in digenous forest precursor communities, grassland and fynbos in coastal vegetation near Port Elizabeth.
Acacia saligna ( Figure 7E) had a similar distribution to A. cyclops, being restricted to the coastal belt. However, it was only abundant in the Port Elizabeth area in moun tain fynbos on the lower slopes of the Vanstadensberg and Elandsberg and in dune vegetation surrounding the airport. It is spreading rapidly in the Grahamstown area and needs to be closely watched (A. Jacot Guillarmod pers. comm.). (Figure 81), the sweet brier rose of Europe and Britain, was brought to the eastern Cape by English settlers during the 1820's and shortly afterwards (Palmer 1985). By 1937 it was reported to be a nuisance in the mountainous parts of Barkly East and a possible threat to the indigenous vegetation (National Herbarium, Pretoria). Like many other members of the family Rosaceae it appears to require low winter temperatures to terminate seed dormancy. Its present distribution as a naturalized plant in southern Africa is largely confined to the mountainous districts of Lesotho, Natal, Orange Free State and northeastern Cape (Jacot Guillarmod 1971; National Herbarium, Pretoria). These regions experience the highest frequencies of below-freezing minimum tem peratures in southern Africa (Tyson 1986).

Rosa eglanteria
This survey showed Rosa eglanteria to be heavily inva sive in the districts of Barkly East, Rhodes, Naudesnek, Rossouw and Jamestown. It is said to be spreading rapidly in the Rhodes area and that the fruits are eaten by people, Angora goats and birds (W.A. Steynberg pers. comm.). I predict that Rosa eglanteria will become increasingly abundant and troublesome throughout the high altitude grasslands situated on the Stormberg and Drakensberg plateaus, i.e. from Molteno northeastwards to the Lesotho border. However any control programmes must take into account its possible value as a food plant and a source of revenue for local people. According to Palmer (1985) there is a factory in the eastern Orange Free State which processes the fruits (hips), making a vitamin syrup. In Lesotho every rose area has its annual rosehip holiday when the children pick the fruits to raise money for their schools.
Populus X canescens ( Figure 8C), Salix babylonica ( Figure 8K) and S. fragilis (fide R.D. Meikle) ( Figure 8L) were the most prominent invaders of watercourses after Acacia meamsii. All three species are large (up to 20 m and more in the case of P. x canescens), long-lived and can form pure stands along watercourses. P. x canescens, unlike the other two species, only reproduces by suckering from the roots and in this way can form dense stands. S. babylonica, and apparently S. fragilis, reproduce only vegetatively in southern Africa from severed branches (Henderson 1991c). Fast-flowing watercourses in the mountainous districts favour the propagation of all three species as well as the dispersal of the Salix species. Humans have also assisted the dispersal of Salix species by planting truncheons along riverbanks and in riverbeds. Figure 7B) is potentially the most important invader of watercourses in the Grassland Biome of the eastern Cape. This judgement is based on its inva siveness in the grasslands of the Transvaal (Henderson & Musil 1984), Natal (Henderson 1989), Orange Free State (Henderson 1991a) and Lesotho (Talukdar 1981). Major factors contributing to its success as a riverine invader are its massive production of long-lived seed and the efficient dispersal of seed along watercourses. Dean et al. (1986) report a seed longevity of 100 years for A. dealbata. Biological control using seed attacking enemies would probably be the most effective method of curtailing the spread of both this species and A. meamsii. Conflicts of interest with the Wattle Industry have halted any research in this direction (H.G. Zimmermann pers. comm.). Figure 1C) and Hakea sericea have invaded mountain fynbos in the eastern Cape. In this survey both species were recorded in the Grahamstown area but only A. longifolia was recorded on the mountains near Port Elizabeth. The National Herbarium in Pretoria has a record of H. sericea dating back to 1976 on the Van Staden's M ountain.

Acacia longifolia (
Biological control pro grammes started in the 1970's (for H. sericea) and in the 1980's (for A. longifolia) offer a means of reducing their vigour and curtailing their spread. Reductions of up to 80% in annual seed production of both species have been recorded (Dennill 1987;Gordon 1990). An indigenous fungus causing gummosis and death in H. sericea is particularly devastating (Morris 1982) and has now been used to produce the world's first mycoherbicide (M.J. Mor ris pers. comm).
Three Pinus species were heavily invasive (i.e. scoring abundance ratings of 5 or more) in parts of the eastern Cape. These were P. pinaster ( Figure 8B). P. halepensis ( Figure 8A) and P. patula. P. radiata was locally abundant. Macdonald & Jarman (1984) ranked P. pinaster, P. radiata and P. halepensis as the fourth, seventh and eighth most important invaders of the Fynbos Biome. P. patula is an important invader of moist montane grasslands in Natal (Macdonald & Jarman 1985) and the Transvaal (Henderson & Musil 1984).
All these pines have winged seeds adapted to wind dispersal. Pinus radiata seed is able to travel up to three kilometres from its source (Richardson & Brown 1986). P. radiata, P. pinaster and P. patula are all reported to regenerate profusely from seed after a fire (Kruger 1977;Richardson & Brown 1986;Wormald 1975). These winddispersed and fire-adapted pines are a particular threat to the mountain fynbos and moist subtropical grassland of the eastern Cape. (Figure 8J), recorded during this survey, and R. phoenicolasius reported by Phillipson (1990) are potentially important invaders in moist subtropical grass land. Both species are well-established near Hogsback in the Amatole Mountains.

Rubus affinis
Several species which are heavily invasive along the coastline of Natal in Acocks's Coastal Forest and Thornveld (Henderson 1989), are also invasive in the eastern Cape at the southern limit of the same veld type. These species are Psidium guajava ( Figure 8F), Lantana camara ( Figure 7J), Solanum mauritianum ( Figure 80) and Cestrum laevigatum. They could become serious invaders within this veld type in the eastern Cape which stretches from the Transkei border to about 50 km south of East London near the Keiskamma River.
Chromolaena odorata, not recorded in this survey, is potentially the most important invader of the stretch of coastline just mentioned. It has been rated as the most important invader in Natal (Macdonald & Jarman 1985) and is largely confined to Acocks's Coastal Forest and Thornveld (Henderson 1989).
Pereskia aculeata, a climbing cactus, is another impor tant invader of coastal forest in Natal (Macdonald & Jarman 1985) and a potentially important invader in the eastern Cape. It was not recorded in this survey but has been reported to be spreading in the Grahamstown and Bathurst areas by Jacot Guillarmod (1988).
Leucaena leucocephala, not recorded in this survey, is a potentially valuable fodder and firewood plant, and is also a potential invader of the coastal lowlands of the eastern Cape. This species is invasive in Natal (Macdonald & Jarman 1985) and is a serious weed in several coun tries in the tropics (Holm et al. 1979). The Department of Agricultural Development has up till now prevented the importation of commercial quantities of seed but it does recognise that Leucaena has much potential and should be exploited (V.D. Wassermann pers. comm.). Certain cultivars should be promoted in specific areas but this should exclude the Hawaiian type because of its prolific seeding. Consideration is being given to the introduction of suitable seed-eating insects with a view to curbing further spread of this species in affected areas (V.D. Wassermann pers. comm.).
Opuntia stricta commonly known in South Africa as the Australian Pest Pear because it reached pest proportions in Australia (Mann 1970) is another potentially important invader. It was seldom recorded during this survey but it could have been overlooked because of its low stature. It is said to be spreading in the Savanna Biome between Alex andria on the coast and Grahamstown (H.G. Zimmermann pers. comm.) It is also an invader of savanna vegetation in the northern Cape (Henderson 1991), Natal (Hender son 1989), Transvaal (pers. observ.) and Namibia (Brown & Gubb 1986). It has invaded an area of approximately 10000 ha south of Skukuza in the Kruger National Park situated in the Transvaal (K. Maggs pers. comm.).
Apart from the riverine invaders already mentioned, only a further four species were recorded as heavily inva sive (i.e. scoring abundance ratings of 5 or more) in one or more localities. These were Acacia longifolia in Howison's Poort near Grahamstown; Arundo donax ( Figure 7G) on the coast near East London and in Acacia savanna near Adelaide; Ligustrum sp. in moist subtropical grassland in the Amatole Mountains; and Sesbania punicea ( Figure 8N) on the coast near East London and along the Gamtoos River valley in the Hankey and Patensie Districts. The latter infestations are being cleared with the use of herbicides (H.G. Zimmermann pers. comm.). There is much confidence that a biological control programme, initiated in the 1980's and using three species of introduced weevils, will halt the invasive spread of this plant in South Africa (Hoffmann & Moran 1988).
Species which have not already been discussed and which were heavily invasive in one or more localities in roadside and veld habitats were: Eucalyptus diversicolor on the Elandsberg near Port Elizabeth, and unidentified species of Eucalyptus (possibly relics of a dune stabiliza tion programme) in dune vegetation near Port Elizabeth ( Figure 71); Nicotiana glauca in karoo vegetation near Jansenville ( Figure 7M); Pennisetum sp. in coastal vege tation near Kidd's Beach (East London District), and Ricinus communis in coastal vegetation near Alexandria and in the Gamtoos River valley near Patensie ( Figure 8G).
Ricinus communis has generally been regarded as an introduced species in southern Africa possibly from else where in Africa. However, seeds in excess of 1 200 years old have been discovered in archaeological diggings in the Baviaanskloof near Patensie (Brink 1988). This evidence suggests that, if indeed introduced, primitive huntergatherers were the agents (Brink 1988). This is in sharp contrast to the majority of our alien weeds which have been introduced since the colonization of the Cape 300 years ago (Brink 1988;Wells et al. 1986).
Fifteen species were locally common in one or more localities. These were Acacia melanoxylon, Pinus radiata and Solanum mauritianum in the Amatole Mountains near Stutterheim (all three spp.) and Hogsback (A. melanoxylon)', Robinia pseudoacacia ( Figure 8H) (water courses), Populus cf. nigra ( Figure 8D) (watercourses), Agave americana ( Figure 7F) (watercourses) and Prunus persica ( Figure 8E) in temperate grassland; Agave ameri cana and Trichocereus cf. spachianus in arid savanna in the Jansenville District; Airiplex cf. nummularia ( Figure  7H) and Tamarix cf. ramosissima along watercourses in the karroid western parts; Nerium oleander ( Figure 7L) along the Baviaanskloof River; Casuarina cunninghamiana and Phytolacca dioica in the Gamtoos River valley in the Hankey and Patensie Districts; Melia azedarach ( Figure 7K) in disturbed vegetation around East London; Opuntia vulgaris in coastal thicket between Port Elizabeth and Alexandria and Eucalyptus spp. (Figure 71) around Grahamstown. Martin & Noel (1960) estimated that between 15 and 20 Eucalyptus spp., as well as hybrids, grow in and around Grahamstown. It was not possible to say how many were cultivated only and how many were naturalized.
Agave americana appeared to be spreading from seed in the Kamferspoort and surrounding areas in the Grootrivierberge southwest of Jansenville. Several scattered plants were seen growing in high rocky clefts far from any planted specimens. Large plants with copious seed were seen on the plains below the mountains. This was an un usual sighting since A. americana usually spreads only very locally by suckering (pers. obs.). It may also be capa ble of limited spread from bulbils (small plants produced in the axils of the inflorescence).

Relation of invasion to environmental factors
'From historical data it is clear that vast retrogressive and even radical changes have taken place in the indigenous vegetation of the eastern Cape' (Roux & Van der Vyver 1988). These changes have occurred largely since the settlement of European farmers in this region in about 1770 (Jacot Guillarmod 1988). The deterioration of the indigenous vegetation has been associated with overgrazing, poor management practices, bush-clearing and alien plant invasion (Lubke et al. 1986;Roux & Van der Vyver 1988;Teague 1988).
Already by 1776 there were reports that the grazing had started to deteriorate rapidly after only seven or eight years of settlement with cattle in the Camdeboo region near Graaff-Reinet (Jacot Guillarmod 1988). It is in the same region that the prickly pear Opuntia ficus-indica was first introduced to the eastern Cape and in which it became a serious problem (MacDonald 1891).
While degradation of the indigenous vegetation opened the way for alien plant invasion, there were other factors which influenced the success of individual species. The successful spread and invasion of large areas by Opuntia ficus-indica and O. aurantiaca can be largely attributed to their adaptability to the prevailing climatic conditions, their efficient dispersal mechanisms and to the absence of natural predators. MacDonald (1891) reported that the seeds of O. ficus-indica were spread in the excreta of humans, baboons, birds, cattle, sheep and goats. Even the Addo elephants, before they were fenced in at the Addo Elephant Park, were reported to eat the fruit of O. ficusindica (Archibald 1955). Today the elephants have virtually eliminated O. ficus-indica from the Addo Park (Macdonald 1984). O. aurantiaca spreads only vegetatively by detached stem sections. These sections are very spiny and readily attach themselves to animals, clothing, shoes and even vehicles. Stem sections of both O. ficus-indica and O. aurantiaca are dispersed by water.
The absence of natural predators appears to have been one of the most important factors in the successful invasion of Opuntia ficus-indica, O. aurantiaca and O. vulgaris in South Africa. This was demonstrated by the dramatic destruction of dense populations of these species, including the almost complete eradication of O. vulgaris following the introduction of their natural insect herbivores (Zimmermann et al. 1986). Zimmermann et al. (1986) con clude that insect herbivores are also likely to play an important role in determining the abundance and distri bution of other alien plant species in South Africa.
The success of some invasive species in the eastern Cape has no doubt been aided by their establishment in large plantations. This certainly seems to be the case with species of Pinus and Acacia. Notable species which have become invasive are Pinus pinaster, P. halepensis and Acacia meamsii, all of which have been cultivated commercially for their timber and in the case of A. meamsii, for the tannin in its bark. P. halepensis, A. cyclops, A. saligna and A. pycnantha were used for driftsand reclamation at Port Elizabeth between 1893 and 1897 .
Water, or the lack thereof, has possibly been the most important abiotic factor influencing alien plant invasion in the eastern Cape. In terrestrial habitats, most invasion in terms of species diversity and abundance of invaders was recorded in the wetter eastern parts. With the excep tion of invasion by a few drought-adapted species, most invasion of the arid central and western interior has been noted only along watercourses.
Watercourses have enabled the long-range dispersal of many species including those which otherwise would be relatively immobile, such as Acacia meamsii, A. dealbata, Sesbania punicea and Ricinus communis. Salix babylonica and 5. fragilis are restricted to watercourses and depend on flowing water for their vegetative dispersal.
Invaders which have successfully invaded fynbos (a fireadapted vegetation type) have various adaptations which enable them to survive periodic high intensity fires. These adaptations include serotiny (seeds held in heat resistant cones) in Hakea sericea and Pinus pinaster, and firestimulated seed germination in Acacia longifolia and A. meamsii.

SOME IDEAS FOR THE FUTURE
Alien plant invasion is likely to increase in all parts of the eastern Cape and particularly in the wetter eastern parts from sea level to an altitude of about 1 300 m. The subregions and their indigenous vegetation types which are most at risk are the coastal belt between the Kei and Keiskamma Rivers (coastal 'forest'), the coastal mountain ranges (mountain fynbos) and southern interior mountain ranges extending from Stutterheim to Somerset East (moist subtropical grassland).
Many invasive species are so well established that their eradication is probably not possible nor feasible. Efforts should however be made to contain their spread and prevent their invasion of new sites. Control programmes should take into account the species complexes which occur in all vegetation categories. The removal of one problem species could simply open the way for other problem species.
Urgent attention should be given to the control, or if possible, the eradication of potentially important invaders which are relatively scarce at this stage. These include Opuntia stricta and Pereskia aculeata. Steps should be taken to prevent the spread of Leucaena leucocephala from plantations. Chromolaena odorata, not yet recorded in the eastern Cape, is a potentially serious invader of the coastal belt between the Kei and Keiskamma Rivers. This species must not be allowed to establish itself in the eastern Cape.
Some research priorities suggested are the hydrological impacts of alien plant invaders along watercourses and in mountain catchment areas; the breeding of sterile cultivars of useful but invasive species and methods for the control and utilization of invader species.