First record of the North African Launaea arborescens in southern Africa

The first record of the non-native, naturalised Launaea arborescens (Batt.) Murb. in the Namib Desert raised questions of its origin and whether or not it could pose a threat to the indigenous vegetation. The North African plant was introduced in a forestry nursery in the Kuiseb Delta, and some individuals were also planted outside the nursery in the early 1970s. They have maintained a likely viable population for nearly 50 years but have so far not been observed elsewhere and thus appear not to be spreading.


Introduction
Invasive alien plants displace native vegetation and change ecosystems all around the world (Chapin et al. 2000). Because of their harsh environments, deserts have so far been spared the deluge of invading plants typically encountered on islands and other ecosystems that are easy to invade (e.g. Duarte et al. 2005;Davies et al. 2011). However, even deserts are not immune to these invasions, and particularly habitats that are moister and receive water regularly such as ephemeral rivers, seeps and fountains can provide suitable habitats for invaders. All western-flowing rivers of the Namib Desert, for example, are locally invaded by numerous exotic species such as Datura, Prosopis, Nicotiana and Ricinus species (e.g. Boyer & Boyer 1989;Auala et al. 2014). The level of infestation varies and is currently localised, but the invasives are nevertheless there. The densest infestations are usually in the vicinity and downstream of settlements where human activities such as woodcutting and overgrazing have disturbed the natural balance.
However, exotic, non-native plants also turn up in unexpected places -for example, in the northwest corner of the Kuiseb Delta, far from settlements and in an area hardly ever touched by the infrequent Kuiseb River floods. This was not recognised for decades, though the spiky desert shrub Launaea arborescens (Batt.) Murb. was found here recently and this raised numerous questions: How did it get there? As this plant is not known in Namibia, has it been reported elsewhere in southern Africa and in similar habitats? Has it established a viable population, and does it pose a threat as an invader? Whether Launaea arborescens has established a viable population in the Kuiseb Delta was investigated by posing two subquestions: What is the extent and size of the population, and what is the size distribution of the plants. These questions are addressed and background information on the plant is presented in this article.
First record of the North African Launaea arborescens in southern Africa

The study area
The study area is situated in the central Namib Desert, in the Erongo Region of Namibia, just south of Walvis Bay (Figure 1). The central Namib lies between the ephemeral Ugab and Kuiseb rivers, and is bounded by the Atlantic Ocean in the west and the escarpment in the east. It falls into the Desert Biome (Irish 1994;Rutherford & Westfall 1994). The coastal zone at Walvis Bay lies within a 'cool desert' region of Namibia, an environment influenced by the South Atlantic Anticyclone, the cold, northward-flowing Benguela current and the divergence of the southeast trade winds along the coast. According to Mendelsohn et al. (2002), average daily temperatures vary between a minimum of 10°C in the coldest month and a maximum of 24°C in the warmest month in the area, although temperatures as low as 2°C have been recorded (Jürgens et al. 2013). Temperatures are variable both daily and seasonally, with the highest temperatures recorded during 'berg wind' episodes, when cold air from the interior flows towards the coast and is heated by compression (catabatic wind), resulting in temperatures of up to 40°C or more. Southerly, westerly, and southwesterly winds are prevalent, and are usually strongest between late afternoon and early evening (Mendelsohn et al. 2002).
Rainfall in the Namib Desert is highly variable, unpredictable and patchy. It varies from 0 to approximately 100 mm per annum, increasing from west to east. In the west, where precipitation from rain is lowest, fog is carried inland by wind passing over the cold Benguela current of the Atlantic Ocean. It is a vital source of moisture for many desert organisms. Walvis Bay has a mean annual rainfall of 13.5 mm, with most rain falling in summer between January and April, and the wettest month being March when about 50% of annual rainfall is recorded (Atlas of Namibia Team 2022).
Coastal dune hummocks are the prevalent habitat in the study area. They are important components of coastal ecosystems and are formed mainly by the shrub Salsola nollothensis Aellen, which traps windblown sand and results in the vertical formation of dune hummocks. These sand-stabilising hummocks are ecologically important, providing shelter, forage and habitats for the local fauna and flora.

The study plant
Launaea arborescens is growing in a dune hummock area, where hummocks are of medium size (1-2 m high). Most of the other plants in this area grow on these hummocks, but also in valleys in between. However, they are absent in valleys where the ground seems to be wet regularly (i.e. closer to the water table and likely saline conditions).
Vegetation cover in the dune hummocks averages approximately 20% and includes, in addition to L. arborescens, the shrubs Salsola nollothensis and Lycium tetrandrum, as well as the herb Crotalaria colorata, the grass Odyssea paucinervis and the sedge Scirpus dioicus. Phragmites australis reed beds adjoin the area to the west. A few isolated trees such as Tamarix usneoides, the exotic Acacia cyclops, Myoporum serratum and Eucalyptus camaldulensis appear to have escaped from the forestry nursery at Wortel and are also found in the hummock area nearby.
The plant was originally identified by recognition of its similarity in flower structure to its closest indigenous relatives in Namibia, namely Launaea intybacea (Jacq.) Beauverd and Lactuca inermis Forssk. They all belong

A B
to the Tribe Cichorieae within the daisy family (Asteraceae), which is characterised by the presence of milky latex and (in most species) bisexual ligulate florets with five teeth at the tips (i.e., strap-shaped 'petals' with toothed tips. This tribe includes many edible, medicinal and weedy species, in genera such as Lactuca (lettuce), Cichorium (chicory/endive), Sonchus (milk thistle) and Taraxacum

Review of information
To establish whether this plant is recorded for the first time in southern Africa, global biodiversity databases

Extent of the population
The extent of the population was established by mapping the occurrence of the plant in the field. The locality of all plants encountered was recorded with a geographic positioning device (GPS). These were then plotted, and the apparent centre determined, which appeared to be just outside the entrance to the nursery. From this centre transects were walked in all compass directions and the distribution boundary recorded. Seven transects of 150-320 m length were walked. The plants were counted along the transects and in the areas between these transects, which were also searched thoroughly. The nursery area was also surveyed (Figure 1B). The GPS positions of recorded individuals and groups of individuals were plotted and the area of occupancy (i.e., the habitat where the plant grows) mapped ( Figure 1B).

Size distribution of the population
The dimensions of 51 haphazardly selected plants (by walking backwards and throwing a bottle over the shoulder and then selecting the nearest plant to where the bottle landed) were measured. At each plant, height and diameter of the plant was measured three times, resulting in six measurements for each individual. Repeated measurements were used to accommodate the irregularity of plant shape and the irregular surface. Height measurements were averaged per individual and plotted in a histogram ( Figure 4). The data are provided in the supplementary (Appendix 1).

Results
The species' history in Namibia   Considering that this plant was introduced some 50 years ago and still likely maintains a sizable population, it is clear that the plant is well adapted to the conditions in the Kuiseb Delta ( Figure 3). However, the question arises whether these are still the original plants that were planted in trials, or if the species has established a self-sustaining, or perhaps even an invading population. Unfortunately, Mr le Roux could not enlighten us on the exact number of plants per species used in the trials, nor the precise localities where these were planted in the study area. However, this is not surprising after such a long time.

Extent of population
The plants are growing to the west and north of the former forestry nursery at Wortel in three distinct local populations. The three mapped areas amount to approximately 3.49 ha. Some 360 individuals were counted. Although the survey aimed to count all individuals, there are possibly more plants as some may not have been discovered in the uneven terrain and between the other plants present. However, the survey provides a reasonably good estimate.

Size distribution of the population
The extent and size of the population contribute to determining whether a population can be considered self-sustaining and therefore viable. Isolated individuals of a species far from other individuals of the same species can hardly present a viable population, particularly if these are out-crossers and therefore rely on genetic exchange with other individuals (Harper 1977;Drew & Kaufman 2013). The size distribution of the plants provides an indication of the population status ( Figure  4). A healthy population contains individuals of all size classes, usually with the majority in the medium range of sizes measured (Harper 1977;Barbour et al. 1987).
To facilitate incorporation of this new record into global databases (e.g., the Global Naturalized Alien Flora database (GloNAF) (Pyšek et al. 2017), we provide the available information on this species in Appendix 2. The parameters proposed by Zengeya and Wilson (2020) were used to structure the data as well as the criteria provided by the IUCN (2020), with adaptations to the terminology suggested by Groom et al. (2019).

Other observations
Launaea arborescens was not the only plant tested for dune stabilisation in Walvis Bay (Le Roux 1974) and other non-native plants in the study area were also noted. Outside the nursery area a few individuals of Acacia cyclops, Eucalyptus camaldulensis and Myoporum serratum were observed. Whether these had been planted originally at their current position or escaped from the nursery and established new plants, could not be established. There were also more trials along the road to Walvis Bay where various non-native Atriplex species had been planted. These have persisted for some 50 years without management intervention.

Discussion
Does Launaea arborescens form a viable population?
The current area of occupancy of the plant of over 3 ha and the well over 300 individuals suggest that this may be a viable population. Also, a bell-shaped size-class distribution of plant sizes is considered indicative of a healthy population in long-lived species such as trees and shrubs (Cousins et al. 2014). L. arborescens size class distribution in the Kuiseb Delta is close to this bellshaped slope with a peak in the medium-sized height class ( Figure 4).
Furthermore, the fact that this plant has been present in the Kuiseb Delta for several decades without irrigation or any other management, suggests that these now form a viable population. Although the plants were initially introduced and irrigated (Le Roux 1974), no maintenance has taken place for at least 40 years. This supports the notion that they may have established a self-sustaining population.

Is there a threat of invasion?
Because no detail was available regarding the number of plants initially planted on a trial basis, or exactly in which areas, it is impossible to establish whether the population has grown or spread. The question of whether the plant has invaded natural areas and replaced indigenous species, or whether they have only maintained their presence in the area where they were initially planted, cannot be answered. Although the plant has not been found in the broader area, monitoring of the population is required over several years (likely decades) to establish whether the plant does form a threat. This study provides a baseline against which future monitoring could be evaluated. However, the fact that the plant has to date also not been found further afield, suggests that it has not spread. However, there have been disturbances in the area since the field survey in the form of widening the two-spoor track to the nursery and beyond with a dozer to more than double the original width, which opens new areas to be occupied by the plant. Human disturbances such as the creation of roads often provide convenient pathways for invasive species (e.g. Higgins et al. 1999;Kumschick et al. 2020;Zengeya & Wilson 2020) and monitoring the plant now may be opportune timing.
Regarding the aspect of potential displacement of indigenous species, the dominant native plants in the surrounding area, which Launaea arborescens could have displaced, are Salsola nollothensis and Lycium tetrandrum. They occupy the same habitat and possibly the same niche, both being shrubs and in the case of L. tetrandrum, with a very similar growth form. The Namib-endemic Crotalaria colorata is a multi-seasonal herb and occupies a different niche. It grows in between the L. arborescens shrubs and is unlikely to have been displaced by L. arborescens.
The effect this exotic plant has on the environment (e.g., soils, water availability) and animals associated with (potentially displaced) indigenous plants in the area, is unknown and may deserve study. It is well known that exotic plants have the potential to alter the environment in many ways (e.g., Chapin et al. 2000) and this aspect should thus not be dismissed. In terms of potential pathways, like many Asteraceae, the diaspores are wind dispersed and this could potentially enable the plants to spread to new habitats.
At present monitoring rather than eradiation is proposed as management. As the risk of the plant spreading is currently considered low, the plant presents an interesting historic record of past forestry management practices and should thus not be eradicated, unless it can be shown that it does spread and presents a threat to the natural vegetation. Also the number of plants and extent of area occupied are small and thus eradication would be easy, if necessary.

Conclusions
It has taken 50 years for some 360 plants to establish and maintain a population in a very localised (3.49 ha) area and Launaea arborescens has so far not been found anywhere else in the Kuiseb River or other rivers or habitats in the Namib, or even southern Africa.
At present we consider the occurrence of the North African Launaea arborescens in the Kuiseb Delta an interesting record and testimony of the efforts that were made some 50 years ago to find solutions to the problem of sand intrusion into Walvis Bay. However, we encourage people to look out for this plant and report any further distribution records. Should plants appear in places other than near the Wortel nursery or former dune stabilisation trial sites, further study regarding the plants' invasive potential is recommended.

Supplementary Material
Appendix 1 Position and dimension measurements (in cm) of Launaea arborescens individuals in the Kuiseb Delta in Namibia (LAT S= latitude south, LON E= longitude east, H1-H3= height measurements in cm, L1-L3= length measurements in cm).