Risk analysis of alien grasses occurring in South Africa

Thesis (MSc)--Stellenbosch University, 2018.

ENGLISH ABSTRACT: Alien grasses have caused major impacts in their introduced ranges, including transforming natural ecosystems and reducing agricultural yields. This is clearly of concern for South Africa. However, alien grass impacts in South Africa are largely unknown. This makes prioritising them for management difficult. In this thesis, I investigated the negative environmental and socio-economic impacts of 58 alien grasses occurring in South Africa from 352 published literature sources, the mechanisms through which they cause impacts, and the magnitudes of those impacts across different habitats and regions. Through this assessment, I ranked alien grasses based on their maximum recorded impact. Cortaderia sellonoana had the highest overall impact score, followed by Arundo donax, Avena fatua, Elymus repens, and Festuca arundinacea. As with other plant groups, alien grasses cause the most impacts through competition with native species (72% of species) and agricultural crops (57% of species). There was variation in impact magnitudes between regions, and among habitat types, however, these differences were not statistically significant. I also found no correlation between impacts recorded from elsewhere and those recorded in South Africa. I then looked at these impacts and risks in South Africa using six alien Paspalum species as a test case. I used species distribution modelling to determine the extent of their potential distribution and a risk analysis framework to determine their overall risk in the country. The species distribution models indicate that P. dilatatum has higher relative occurrence rate in relation to other alien Paspalum species in South Africa, and P. nutans has the lowest relative occurrence rate. The environmental variable which influenced the relative occurrence rate of most Paspalum species was annual mean temperature and was the most contributing variable of three species: P. dilatatum, P. nutans, and P. urvillei. Temperature seasonality was the main environmental variable for P. quadrifarium, while the relative occurrence rate of P. virgatum and P. nutans was best predicted by annual precipitation. The risk analyses show that P. dilatatum, P. notatum and P. quadrifarium have high risk with medium management feasibility, and the other three (P. nutans, P. urvillei, and P. virgatum) are low risk in South Africa. Finally, based on the results of the risk analysis and the distribution models, I conducted a field assessment of the impacts of P. quadrifarium on the local plant community. I compared sites where P. quadrifarium currently invades, to those that had been invaded and subsequently cleared of the species and to those that had not been invaded. I found that P. quadrifarium altered the plant species composition and reduced their abundance, i.e. it appears to impact plant communities through competition. I also found no regrowth of P. quadrifarium in the site where it was cleared mechanically by uprooting the whole plant. However, there was evidence of secondary invasions by other alien plants in the site, highlighting the need for active restoration. The results of this thesis are relevant for the management of alien grasses in South Africa as they can be used to motivate for their prioritisation and regulation in the country.

AFRIKAANSE OPSOMMING: Uitheemse grasse het ‘n groot impak in hul ingedringde reekse veroorsaak, insluitend die omskepping van natuurlike ekosisteme asook die vermindering van landbouopbrengste. Dit is duidelik kommerwekkend vir Suid-Afrika. Die impak van uitheemse grasse in Suid-Afrika is egter grootliks onbekend, en maak dit moeilik om die bestuur daarvan te prioritiseer. In hierdie proefskrif het ek die negatiewe omgewings- en sosio-ekonomiese impakte van 58 uitheemse grasse in Suid-Afrika ondersoek uit 352 gepubliseerde literatuurbronne, die meganismes waardeur hulle impak veroorsaak en die omvang van die impakte oor verskillende habitatte en streke. Deur hierdie assessering het ek uitheemse grasse gegrond op hul maksimum aangetekende impak. Cortaderia sellonoana het die hoogste algehele impak telling, gevolg deur Arundo donax, Avena fatua, Elymus repens en Festuca arundinacea. Soos met ander plantegroepe, veroorsaak uitheemse grasse die meeste impak deur die kompetisie met inheemse spesies (72% van spesies) en landbougewasse (57% van spesies). Daar was variasie in impakgroottes tussen streke, en onder habitatsoorte was hierdie verskille egter nie statisties betekenisvol nie. Ek het ook geen verband gevind tussen impakte wat elders aangeteken is en dié wat in Suid-Afrika aangeteken is nie. Ek het toe na hierdie impakte en risiko's in Suid-Afrika gekyk deur ses uitheemse Paspalum spesies te gebruik as 'n toetssaak. Ek het spesieverspreidingsmodellering gebruik om die omvang van hul potensiële verspreiding te bepaal, en 'n risiko-analise raamwerk om hul algehele risiko in die land te bepaal. Die spesies verspreidingsmodelle dui daarop dat P. dilatatum 'n hoër relatiewe voorkoms-koers het in verhouding tot ander uitheemse Paspalum spesies in Suid-Afrika, en P. nutans het die laagste relatiewe voorkoms-koers. Die omgewingsveranderlike wat die relatiewe voorkoms van die meeste Paspalum spesies beïnvloed het, was jaarlikse gemiddelde temperatuur en was ook die mees bydraende veranderlike van drie spesies: P. dilatatum, P. nutans, en P. urvillei. Temperatuur seisoenaliteit was die belangrikste omgewingsveranderlike vir P. quadrifarium, terwyl die relatiewe voorkoms van P. virgatum en P. nutans die beste voorspel is deur jaarlikse neerslag. Die risiko ontledings toon dat P. dilatatum, P. notatum en P. quadrifarium 'n hoë risiko het met medium bestuur haalbaarheid, en die ander drie (P. nutans, P. urvillei, en P. virgatum) is lae risiko in Suid-Afrika. Ten slotte, op grond van die resultate van die risiko-analise en die verspreidingsmodelle, het ek 'n veldassessering gedoen rakende die impak van P. quadrifarium op die plaaslike plantgemeenskap. Ek vergelyk plekke waar P. quadrifarium tans indring, met plekke wat reeds ingedring is (en daarna van indringerspesies verwyder is), en ook met plekke wat nog nie ingedring is nie. Ek het bevind dat P. quadrifarium die plantspesiesamestelling verander het en hul oorvloed verminder het, met ander woorde, dit wil voorkom asof dit plantegroei beïnvloed deur mededinging. Ek het ook geen hergroei van P. quadrifarium gevind op die terrein waar dit meganies skoongemaak is, deur die hele plant te ontwortel nie. Daar was egter bewyse van sekondêre indringing deur ander uitheemse plante in die terrein, wat die behoefte aan aktiewe herstel beklemtoon. Die uitslae van hierdie proefskrif is van belang vir die bestuur van uitheemse grasse in Suid-Afrika, aangesien dit kan dien as motivering vir hul prioritisering en regulering in die land.

Masters