The extensive and ongoing invasion by cacti in South Africa has been documented since the 20th century, when Opuntia ficus-indica (L.) Mill. occupied vast tracts of the Karoo (Henderson 2015). A recent global review of Cactaceae invasions identified South Africa as the second most invaded of all countries in respect of this largely New World family, with 35 taxa recorded (Novoa et al. 2014). This study identifies the bioclimatic equivalence of South Africa to the natural range of invading cacti as a factor key to their successful naturalisation on the subcontinent.

In 2008, the Working for Water programme of the South African Department of Environment Affairs funded the South African National Biodiversity Institute Invasive Species Programme (SANBI ISP). The SANBI ISP documents new instances of naturalisation, evaluates taxa on the Species Under Surveillance for Possible Eradication or Containment Targeting (SUSPECT) list, advises on the legal listing of taxa and develops specific eradication plans (Wilson et al. 2013).

Earlier surveys documenting the vegetation of Weenen County (West 1951) and the entire Tugela River basin (Edwards 1967) did not record any members of the Cactaceae, apart from the indigenous Rhipsalis baccifera (J.S. Muell.) Stearn subsp. mauritiana (DC.) Barthlott. However, by 1987 both Opuntia ficus-indica and O. stricta (Haw.) Haw. var. stricta were common in this area, with O. aurantiaca Lindl. recorded there since 1999 (Henderson 2007; Southern African Plant Invaders Atlas [SAPIA] database 2015).

The suitability of the central region of the largest catchment in KwaZulu-Natal (KZN), the Tugela basin, for colonisation by succulents necessitates ongoing surveys for new or extended invasions by cacti, the assessment of environmental risks posed by them and the formulation of possible management approaches. The objectives of this study were to survey the cacti in the central Tugela River and to assess the invasion status of those not previously recorded in the area, as well as to make recommendations for the management of exotic cacti encountered during the survey, based on field observations and the SAPIA database.

Field work in the central Tugela River basin near Weenen was undertaken in early December 2014 to gather cacti distribution data and specimens, particularly of Peniocereus serpentinus. A 40 km section of tertiary road was travelled slowly by car through the topocadastral square 2830 CC, from the R74 main road northward across the Bloukrans River towards the Tugela River. Herbarium specimens were gathered and submitted to the KZN Herbarium.

The invasion stages of three cacti species in the region were categorised according to Blackburn et al. (2011), following the interpretation of Wilson et al. (2014), with management recommendations formulated for all cacti encountered during the survey. A weed risk assessment (Gordon et al. 2010; Pheloung, Williams & Halloy 1999) was undertaken for one introduced cactus species not previously evaluated in South Africa.

Material was gathered under permit number OP 967/2014 issued by Ezemvelo KZN Wildlife.

The first record of Echinopsis oxygona (Link) Zucc. ex Pfeiff. & Otto outside of cultivation in South Africa was confirmed and the first naturalised population of Opuntia microdasys (Lehm.) Pfeiff. in KZN was documented. Four populations of Peniocereus serpentinus (Lag. & Rodr.) N.P.Taylor were encountered, ranging in size from several square metres to 0.4 ha.

The population of O. microdasys detected in the Tugela River basin (Figure 1) represents a significant range extension (± 410 km) into the KZN province (Figure 2). Previously, this species was known in South Africa from 45 sightings (21 localities) in 4 provinces: the Western Cape, Eastern Cape, Gauteng and Limpopo (SAPIA database 2015; Winter, Zimmerman & Mashope 2011). Opuntia microdasys is at the D2 stage of invasion, defined as being a self-sustaining population, outside of cultivation and with plants surviving and reproducing a significant distance from their original point of introduction (Blackburn et al. 2011) (Table 1).

Descriptions of O. microdasys in South Africa have previously been provided by Smith et al. (2011) and Winter et al. (2011). Smith et al. (2011) stated that in South Africa, plants of this species take a number of years to flower and, given that flowers and fruit were observed (Figure 1), the stand is evidently well-established, a notion supported by its expanse of ± 0.025 ha. No cochineal infestation was seen on O. microdasys, despite the presence of these insects on O. stricta var. stricta (Figure 3) and O. aurantiaca plants (Figure 3) growing in close proximity.

A small colony of E. oxygona (Figure 1) was detected adjacent to the O. microdasys population. As this remote site (Figure 4) showed no sign of abandoned habitations or cultivated lands, the manner in which this colony started is unknown. However, based on the size of the multiple stems (> 35 cm), it is evidently well-established.

A brief description of E. oxygona sensu lato has been provided by Hunt, Taylor and Charles (2006), with expanded descriptions of its synonymised elements published elsewhere (Borg 1963; Britton & Rose 1963). The plants of the Tugela River basin possess somewhat glaucous, clustered stems that are subglobular. Later they are short, cylindric-clavate, up to 65 cm high and 10 cm broad, and sprawling with many shoots at the base and along the sides. The 13 to 16 ribs are continuous, broad at the base and with deep furrows between them. The circular areoles are ± 2 cm apart, with whitish wool. Young shoots are distinctly spiny, whilst mature stems are relatively spineless. Flowers open at sunset, are narrow-funnelform, ± 20 cm long and 12 cm in diameter. The inner perianth segments have a spreading growth, and are translucent and rose-coloured. The axils of scales on the flower tube bear long, gray to blackish wool. The filaments and style extend a little beyond the throat, but are shorter than the perianth segments (Figure 1). The genus is reported by Borg (1963) to produce oval, fleshy, hairy berries, at maturity splitting on one side to reveal very small and minutely pitted seeds.

Echinopsis oxygona can be considered to be at the C3 stage of invasion, having a self-sustaining colony that is reproducing vegetatively and possibly sexually (flowering, although no fruit was observed) (Table 1). Following the weed risk assessment model of Pheloung et al. (1999), designed to screen primarily for invasive plant taxa, E. oxygona generates a score of 12, which falls into the reject category. This result is based on the answers to 49 questions that evaluate the plant’s current biogeography, the impact of any undesirable attributes it may have and its ecophysiology (Pheloung et al. 1999). This means that an application for its import into South Africa should be turned down. As a first record for KZN, O. microdasys may still be restricted to a small region in the central Tugela River basin, although it will likely spread in the absence of control interventions, which are suggested in Table 2.

Four P. serpentinus populations, which were over 1 km apart, were detected (Figure 3). These ranged from a few square metres to 0.4 ha in extent. The largest population, of a rectangular, polygon shape, was located at 28°48′24.5″ S, 30°3′12.1″ E. The most extensive infestation was surrounded by small satellite clumps that were less than 50 m away from the main clump. This taxon is now known from a total of seven localities in KZN and two in Limpopo province (Winter et al. 2011). The upright, clambering stems of P. serpentinus were observed to attain heights of up to 4.5 m, supported by the branches of trees through which they were growing, in conditions ranging from full sun to shade.

Given the distances between the known P. serpentinus populations, and that all but one of these had flowering plants, and in one case fruit, they are classified at the D2 invasion stage (Table 1). It is currently unknown what role people play in the dissemination of this species in the Tugela River basin, and its origin. However, given the rectangular shape of the largest population, it may originally have been cultivated as a barrier plant.

Opuntia microdasys (teddy bear cactus) has a wide natural distribution from Texas to southern Mexico. It is considered to be invasive in Australia (United States Department of Agriculture 2015) and South Africa (Smith et al. 2011) and naturalised in the Iberian Peninsula (Sanz-Elorza, Sánchez & Vesperinas 2006). At a national level, O. microdasys is too widespread and established to be considered an eradication target for SANBI ISP and is categorised as 1b (Department of Environmental Affairs 2014b), meaning that its control needs to form part of an invasive species management plan (Department of Environmental Affairs 2014a; Wilson et al. 2013).

According to the definition by Pyšek et al. (2004), O. microdasys can be considered naturalised in KZN as this self-sustaining population has likely existed for at least 10 years without direct human intervention, by recruitment from seed and ramets. Manual extraction and chemical control with herbicides should enable eradication at this site, but if the species is found to be more widespread in KZN, the feasibility of biocontrol should be considered.

Echinopsis oxygona, the Easter lily cactus or sea urchin cactus, is indigenous to Brazil, Paraguay, Uruguay and Argentina (Hunt et al. 2006). It is listed in Brazilian state legislation as Vulnerable in the Rio Grande do Sul region (De Freitas et al. 2010) and is one of several subjects in an ex situ conservation project in Transylvania (Stoie 2007). However, its global Red List status is Least Concern (International Union for the Conservation of Nature 2014).

The only species of Echinopsis Zucc. currently listed in the Alien and Invasive Species list (Department of Environmental Affairs 2014b) is the torch cactus, E. schickendantzii F.A.C.Weber (syn. E. spachiana [Lem.] Friederich & G.D.Rowley), although E. chamaecereus H.Friedrich & Glaetzle has recently been recorded outside of cultivation (SAPIA database 2015). Echinopsis schickendantzii, a category 1b invader, has naturalised extensively in the central dry interior of South Africa, and remains poorly controlled despite being declared an invader in 2001 (Department of Agriculture 2001). Its succulent fruits are believed to be distributed by both mammals and birds (Henderson 2015) and both of these are likely to also be the dispersal agents of E. oxygona.

The Easter lily cactus, E. oxygona, is naturalised in Spain (Laguna-Lombreras et al. 2014), the Galápagos Islands (Randall 2012) and Australia (McFadyen 2012). It was earlier noted (as E. multiplex Pfeiff. & Otto) as having been cultivated within the Kruger National Park and required removal (Foxcroft et al. 2003). In the early 19th century, various Echinopsis species were introduced into Europe (Mottram & Quail 2009), amongst them the ‘pure species’ E. multiplex, E. oxygona, E. eyriesii (Turpin) Pfeiff. & Otto and E. tubiflora (Pfeiff.) Zucc. ex A.Dietr.

Subsequent crossing of these forms resulted in what were perceived to be a multitude of hybrids of indeterminate parents, such that the pure species were considered to have largely disappeared from cultivation (Bertrand & Guillaumin 1957). However, Hunt et al. (2006) subsequently placed E. multiplex, E. eyriesii and E. tubiflora in synonymy under E. oxygona, so circumscribing a much broader species concept under which the central Tugela River basin material is placed. This species has long been a horticultural subject in South Africa (Glen 2004; Pienaar & Smith 2011).

Within the last two years, several reports have been received by SAPIA of round, cushion-type Echinopsis colonies occurring in non-cultivated situations (Figure 4). In the absence of flowers, it remains uncertain whether all these are attributable to E. oxygona, or to various other taxa in this genus of ± 72 species (Hunt et al. 2006). The risk assessment score of 12 for E. oxygona, as determined in this study, leads to a recommendation that this cactus should not be imported into the climatically similar Australia, another cactus invasion hotspot (Novoa et al. 2014).

In considering the wide distribution of the congeneric E. schickendantzii, and observing that cushion-type Echinopsis naturalisations in South Africa occur in four different veld types across three biomes (Mucina & Rutherford 2006; SAPIA database 2015; Figure 4), E. oxygona likely has a much wider potential national distribution. An adequate sample size of distribution records of this species is currently unavailable on the Global Biodiversity Information Facility (2014) for a reliable distribution model (fewer than 20 with co-ordinates).

The lack of detailed native range data is common to many cacti and in view of such gaps, the South African Cactus Working Group recently proposed a ban on ‘any cacti genera that contain invasive species’ (Novoa et al. 2015). This approach was put forward by Novoa et al. (2015) as the most effective way of regulating the movement of invasive species, without curtailing the horticultural industry unnecessarily. It is recommended (Table 2) that E. oxygona be placed on the SUSPECT list to establish whether its formal regulation is necessary.

Peniocereus serpentinus, the serpent cactus or snake cactus, is indigenous to Mexico, in the Sinaloa region and the states between Querétaro and Oaxaca (Arreola et al. 2013; Hunt et al. 2006). It has naturalised in Queensland, eastern Australia (Atlas of Living Australia 2015; McFadyen 2012). This species is cultivated globally as an ornamental plant (Arreola et al. 2013; United States Department of Agriculture 2015) and, under the synonym Nyctocereus serpentinus (Lag. & Rodr.) Britton & Rose, has been grown in South Africa since at least the mid-1960s (Rawé 1966). Individuals characteristically produce long, thick, turnip-like roots that comprise a substantial proportion of their biomass. The etymology of the generic name is a reference to the opulence of the root growth relative to the development of aerial parts. Winter et al. (2011) provide a description of P. serpentinus and a distribution map that remains largely unaltered by the current report. In South Africa, this species is noted as fast-growing under a range of light conditions (Rawé 1966), as observed in the Tugela River basin.

Peniocereus serpentinus is listed as 1b on the most recent national regulations (Department of Environmental Affairs 2014b), and therefore falls outside the ambit of SANBI ISP. However, given the observed tendency of P. serpentinus to spread and its current localised distribution in the central Tugela River basin, urgent action is required, with recommendations for control provided in Table 2.