Thesis (MSc)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: Alien invasive species are a global concern not only threating biodiversity, but also negatively impacting regional economies. Consequently research aimed at understanding the invasion process is crucial for management of invasive alien species to reduce these impacts. Identifying accurate natal ranges is the key first step for efficient alien plant management, especially biological control. Such knowledge may assist in selecting host-specific biological control agents and help prevent non-target impact on native species. A case in point is the genus Cardiospermum of which species have been extensively moved around the globe with two species now being possible widespread invaders (C. grandiflorum and C. halicacabum), and a third species (C. corindum) having significant invasion potential. However, in some regions the native statuses of these species are not clear, hampering management. In order to minimize potential non-target impacts it is prudent to determine the relationships between Cardiospermum taxa in unknown native ranges prior to the release of biological control agents. With this thesis I aim to review available literature on the genus Cardiospermum, determine potential spread to suitable habitats globally for selected taxa, resolve unknown native ranges in southern Africa, and investigate rapid speciation of an endemic Namib Desert species of balloon vine. Through exhaustive searches for available literature on the genus Cardiospermum I report on the biology and ecology of selected species within the genus, with special focus on the most widespread species. Specifically, using species distribution modelling I investigate the potential spread of C. halicacabum, C. corindum and C. grandiflorum globally and assess the accuracy with which this can be determined using known native ranges to predict current alien ranges. Results indicate that these species have significant potential to spread, though highlighting that species distribution modelling over-fit predicted suitable ranges when using native range data alone. To resolve uncertain native ranges of selected species within the genus I reconstructed a dated multi-gene phylogeny and chloroplast haplotype network to investigate relationships within and among Cardiospermum species. I conclude that C. halicacabum is non-native in southern Africa due to polyphyletic relationships among accessions, while C. corindum is native due to natural long distance dispersal as indicated by a single monophyletic southern African clade. The close relationship observed between C. corindum and the African species, C. pechuelii, lead to a population-level genetic investigation of these two taxa. For this, I constructed a multi-gene phylogeny, chloroplast haplotype network and conducted population genetic diversity analyses which all indicated that the African arid adapted taxa, C. pechuelii, endemic to the Namib Desert of Namibia, evolved from C. corindum following long distance dispersal from South America. This study firstly provides insights into the ecology and biology of Cardiospermum, warning against further introduction due to potential spread. Secondly I resolved the unknown native statutes of Cardiospermum species in southern Africa and, lastly, illustrated the origin of African C. pechuelii. My research findings have major implications for on-going biological research against C. grandiflorum in South Africa and cautions strongly against the release of already-identified biological control agents due to their potential non-target impacts. / AFRIKAANSE OPSOMMING: Indringerspesies is 'n wêreldwye probleem as gevolg van hul direkte bedreiging van biodiversiteit asook hul negatiewe impakte op streeksekonomieë. Gevolglik is navorsing aangaande die indringingsproses van uiterste belang om doeltreffende beheer van indringerspesies te verseker. Die bepaling van die inheemse herkoms van indringerpopulasies is die eerste belangrike stap tot doeltreffende beheer, veral biologiese beheer. Hierdie informasie kan help met die identifisering van gasheer-spesifieke biologiese agente en om potensiële nie-teiken impakte op inheemse plant spesies verhoed. ‘n Voorbeeld is ballon rankplante in die genus Cardiospermum, waarvan verskeie spesies wêreldwyd verspreid is, met die gevolg dat minstens twee spesies (C. halicacabum en C. grandiflorum) nou moontlike wydverspreide indringers is en ‘n derde spesie (C. corindum) ook potensiaal toon as ‘n indringerspesie. In sommige streke is die in- of uitheemse statusse van hierdie spesies onbekend en belemmer gevolglik hul doeltreffende beheer. Met hierdie tesis streef ek om die genus Cardiospermum te hersien en potensiële verspreiding van geselekteerde spesies na geskikte habitatte te bepaal, om onbekende inheemse streke binne Suider-Afrika op te los en laastens, om die herkoms van ‘n ware Afrika spesie (C. pechuelii) te ondersoek. Na deeglike ondersoek van beskikbare literatuur doen ek verslag omtrent die biologie en ekologie van spesies binne Cardiospermum, met spesiale fokus op die mees wydverspreidste spesies. Deur gebruik te maak van spesies verspreidings modelle bestudeer ek die potensiaal van verdere verspreiding wêreldwyd vir C. halicacabum, C. corindum en C. grandiflorum en rapporteer ook omtrent die akkuraatheid van hierdie tegniek. Resultate toon dat alle spesies aansienlike potensiaal toon om verder te versprei, maar beklemtoon ook dat hierdie metode inheemse streek data “oor-aanpas”. Om enige onsekerhede aangaande inheemse streke op te los, het ek ‘n gedateerde multi-geen filogenie en ‘n chloroplas haplotipe netwerk opgetrek om die verhoudings binne en tussen geselekteerde Cardiospermum spesies te bepaal. Ek kom tot gevolgtrekking dat C. halicacabum nie inheems in Suider-Afrika is nie, as gevolg van polifiletiese verhoudings, en dat C. corindum inheems is en moontlik ook ‘n voorbeeld van natuurlike langafstand verspreiding, soos aangedui deur ‘n enkele monofiletiese groep. Die naverwante filogetiese verhouding tussen C. corindum en C. pechuelii het daartoe gelei dat ek die herkoms van C. pechuelii in Afrika ondersoek het. ‘n Multi-geen filogenie, chloroplas haplotipe netwerk en populasie genetiese diversiteit analises het aangedui dat C. pechuelii endemies is tot die Namib Woestyn van Namibië, en die produk is van lang afstand verspreiding van C. corindum vanaf Suid-Amerika. Hierdie studie gee eerstens insig tot die ekologie en biologie van Cardiospermum en waarsku teen verdere verspreiding. Tweedens dui my studie ook die opgeloste inheemse streke in Suider-Afrika aan van sekere Cardiospermum spesies, en laastens, illustreer ek die herkoms van C. pehuelii. My navorsing het baie belangrike gevolge vir die huidige biologiese beheer program teen C. grandiflorum in Suid-Afrika en waarsku teen die vrystelling van reeds geïndentifiseerde biologiese beheer agente weens potensiële nie-teiken impakte op inheemse spesies.

This thesis focuses on the guppy, Poecilia reticulata, as an invasive species. Its non-native distribution, the biology behind its success and the reasons for its introduction are examined. A worldwide email survey revealed that the guppy is established in at least 73 countries outside of its native range and that mosquito control schemes and the release of unwanted aquarium fish are the two primary routes of introduction. Knowledge gaps were identified; primarily the scarcity of scientific evidence for negative impacts of guppy introductions and similarly for mosquito control efficacy. Replicated mesocosm experiments demonstrated that female guppies are capable of routinely establishing populations, and that these retain behavioural viability over several generations. The first mesocosm study suggested that founders with very different evolutionary histories were equally good at establishing populations. The second mesocosm study suggested that monandrous females were extremely successful at establishing behaviourally viable populations, with no decline in behavioural variation. The effectiveness of guppies as mosquito control agents was examined in two related foraging experiments. The first study found little evidence for the presence of ‘prey switching’ in guppies, questioning the validity of previous work advocating their introduction to stabilise prey populations. The second study revealed a preference for non-vector mosquito larvae in a two-prey system. However, both mosquito species were consumed equally readily when habitat complexity increased. The presence of conspecifics affected female foraging behaviour. The presence of males reduced the strength of prey preference in the first study, and the presence of conspecifics of either sex removed prey preference in the second. Both demonstrate that multi-prey systems have important implications for the efficacy of poeciliids in biological control. Despite severe demographic bottlenecks, their adaptability and ability to rapidly increase in numbers enable guppies to establish and persist when introduced. Such bottlenecks are typical of introduction scenarios, warning that particular caution should be exerted when introducing this species, or other livebearing fish, to natural water bodies.