Thesis (PhD)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: Worldwide, invasive alien plants (IAPs) alter aspects of invaded ecosystems including geomorphology, above-ground vegetation, soil seed banks and soil nutrient regimes, thereby affecting the long-term stability of ecosystems. In many cases these invasions call for various management interventions, including restoration. Effects of alien plant invaders on native ecosystems are widely acknowledged, but compared to terrestrial ecosystems, riparian habitats are poorly studied. Riparian habitats are inherently rich in biodiversity, but are particularly prone to invasion by IAPs because of their dynamic nature and because these ecosystems are affected by many anthropogenic activities. This enhances the proliferation of IAPs, especially trees such as Tamarix spp. (e.g. in the southwestern United States), Salix spp. (e.g. in Australia), and Australian Acacia spp. and Eucalyptus camaldulensis (in South Africa). Initiatives such as the Working for Water (WfW) programme in South Africa that are focusing on clearing IAPs, particularly in riparian zones, have reported much success in the short term, but ecosystem recovery remains limited. There is a poor understanding of the nature of the impacts of invasive trees, and of the opportunities that exist for ecosystem restoration. The objective of my thesis was to investigate the effects of the widespread invasions of Eucalyptus camaldulensis in riparian zones. Methodological constraints have been identified as one reason for limited success of restoration projects. Many studies examining the effects of IAPs used a comparative approach of uninvaded (reference) and invaded sites. However, this approach often fails to separate cause from effect. An experimental removal approach has been suggested as an alternative, but it is often inappropriate for large invasive trees owing to time and cost constraints. I used a comparative gradient approach, involving sampling along a continuum from uninvaded sites, and varying E. camaldulensis invasion densities categorised as “lightly”, “moderately” and “heavily” invaded (hereafter the gradient of invasion), each with four replicates (n = 16). This approach enabled me to detect trends that could be reliably attributed to invasion. Furthermore, my study attempted to provide a comprehensive understanding of the effects of IAPs in invaded ecosystems by considering various components of the ecosystem and integrating their feedback relationships to establish the net effects on ecosystem processes. The riparian zone is a unique ecosystem which provides an opportunity for such a multifaceted study examining and integrating the effects of invasive trees on riparian geomorphology, above-ground vegetation, soil seed banks, and soil properties. I studied the effects of E. camaldulensis invasion in riparian zones along the Berg River in the Mediterranean-type climate zone of the Western Cape, South Africa. I asked the questions 1) Are environmental conditions (i.e. the riparian environment created by geomorphological processes) altered by Eucalyptus invasion and if so, are these conditions less suitable for recruitment of indigenous species? 2. Has Eucalyptus invasion altered the composition of indigenous species (potentially limiting post-clearing propagule supply)? 3: How does Eucalyptus invasion change the composition of the riparian soil seed bank; and its potential to re-initiate ecosystem recovery? 4: Do soil properties (physical and chemical) change with Eucalyptus invasion? Data for all four questions were collected in the same sites to enable cross comparisons. For the first question, I employed photogrammetry techniques in a GIS and remote sensing environment to analyse repeated aerial photographs (1938 -2010) to reconstruct the invasion history and riparian geomorphology evolutionary dynamics. For the second and third questions, I compared richness, diversity, evenness and composition of resident above-ground and seed bank vegetation between uninvaded sites and the gradient of invaded sites. For the fourth question, I compared a set of variables describing physico-chemical properties in uninvaded sites to the gradient of invaded sites seasonally. Results of the geomorphology study showed that the riparian zone has been very dynamic over the 70 years, with a 13.5% net decrease in area. There was evidence of channel narrowing, riverbank steepening and river bed incision in areas that were densely invaded. No significant trends were detected in hydrometeorological data. The above-ground vegetation study revealed that species richness, diversity and structural attributes (e.g. height, relative cover and mean basal diameter) of native species decreased consistently along the invasion gradient. Invasion also altered native and alien plant species composition, both of which showed a high preference for lightly invaded sites. The seed bank study revealed that E. camaldulensis invasion had no significant effect on total and native species richness, diversity and evenness, however the effects were bigger on native than alien species. Alien species density was significantly higher than native species density. Invasion also influenced the composition of the native seed bank. However, native seed banks were more diverse than above-ground plant assemblages with some native fynbos species occurring only in the seed bank. In the soil study, I found that soil pH levels were significantly lower in invaded sites than in uninvaded sites in all seasons. Litter cover and thickness increased significantly with invasion intensity. Soil moisture decreased consistently with invasion intensity while temperature increased in winter and spring. Total macro, micro and available nutrients did not vary significantly along the invasion gradient (p > 0.05), but exchangeable cation content was significantly higher in uninvaded than in invaded sites, especially in winter and spring. Narrowing of river channel and steepening of river banks reduces the suitability of the riparian zone to perform certain ecological functions such as soil seed storage and providing space for vegetation to grow. This reduces opportunities for seedling recruitment and consequently reduces species diversity. Conversely, the reduction of species diversity in above-ground vegetation reduces seed input into the soil leading to depauperate soil-stored seed banks. The abundance of seeds of Acacia mearnsii in the soil-stored seed bank poses a threat of secondary invasions post clearing of E. camaldulensis if this is adopted as a management action. The study was successful in addressing the stated objectives. The transition from a native-dominated riparian plant community to a monoculture of E. camaldulensis over c. 50 years has resulted in marked changes to riparian geomorphology and above-ground vegetation, but has yet to radically change the total soil seed bank and soil nutrient concentrations. The findings offer support to the WfW clearing initiatives and show that potential for recovery of native species after removal of the invasive eucalypts still exists. This knowledge not only enhances effectiveness of restoration and long term management of riparian ecosystems, but also advances the field of riparian ecology and restoration. Further experimental work is needed to establish the mechanisms responsible for the changes associated with Eucalyptus invasion. The contribution of agriculture in shaping channel morphology also needs to be investigated. / AFRIKAANSE OPSOMMING: Die langtermyn stabiliteit van ekosisteme wêreldwyd word bedreig deur uitheemse indringer plante (UIPs) wat, onder andere, die geomorfologie, bo-grondse plantegroei, onder-grondse saadbanke en voedingstofstatus van ekosisteme kan verander. In baie gevalle word bestuursoperasies vir indringer plante vereis, en sluit onder andere herstel (restorasie) in. Die invloed van UIPs op inheemse ekosisteme word wyd erken, maar kennis oor oewerbank-habitatte, in vergeleke met ander terrestriële ekosisteme, skiet steeds tekort. Oewerbank-habitatte is inherent ryk aan biodiversiteit maar uiters sensitief vir indringing deur UIPs. Dit is weens die dinamiese aard van hierdie ekosisteme asook vele menslike aktiwiteite in hierdie tipe habitat. Dit dra by tot die vermeerdering van UIPs, veral bome wat byvoorbeeld insluit Tamarix (tamariske) (bv. in die suid-wes van die Verenigde State van Amerika), Salix (wilger) (in bv. Australië) en die Australiese Acacia (wattel) en Eucalyptus spesies (bloekom) in Suid Afrika. Inisiatiewe soos die Werk vir Water (WvW) program in Suid Afrika fokus juis daarop om UIPs vanuit oewerbank areas te verwyder, en alhoewel groot sukses in die kort termyn behaal is, is die herstel van ekosisteme steeds beperk. Daar is dus min inligting beskikbaar oor die impak van indringer bome en gevolglik ook oor geleenthede vir die herstel van hierdie ekosisteme. My tesis ondersoek die impak van die wydverspreide voorkoms van die indringer, Eucalyptus camaldulensis, in oewerbank-omgewings. Beperkte metodes is tans beskikbaar om die impak van indringer plante te ondersoek, en dit is geïdentifiseer as een van die oorsake vir die beperkte sukses van herstelprojekte. Baie studies ondersoek die effekte van UIPs deur natuurlike areas sonder enige indringer plante te vergelyk met areas waar UIPs voorkom. Hierdie tipe benadering kan ongelukkig, in meeste gevalle, nie onderskei tussen die oorsaak en die impak nie. ʼn Alternatiewe voorstel is om UIPs eksperimenteel te verwyder, maar is ongelukkig nie van toepassing op groot bome nie weens die tyd en kostes verbonde hieraan. Ek het gebruik gemaak van ʼn vergelykende-gradiënt benadering wat steekproefnemings langs ʼn geleidelike kontinuum van natuurlike areas sonder enige indringer plante na areas wat effens gematig, en erg ingedring is (hiernaas bekend as die indringer-gradiënt) deur die indringer, E. camaldulensis. Die steekproefnemings van elke kategorie van indringing (geen, effens, gematig en erg) is vier keer herhaal (n = 16). Hierdie benadering het my toegelaat om neigings te identifiseer wat met sekerheid toegeskryf kan word aan die teenwoordigheid van indringers. My studie het ook verder gemik om ʼn omvattende beskrywing van die impak van UIPs in ekosisteme met indringers te verskaf deur verskeie komponente van die ekosisteem, sowel as hul terugvoer-verhoudinge, in ag te neem, en sodoende die totale effek op ekosisteem prosesse te bepaal. ʼn Oewerbank is ʼn unieke ekosisteem wat ʼn geleentheid bied vir ʼn veelsydige studie wat die impakte van indringer bome op oewerbank-geomorfologie, bo-grond plantegroei, grond saadbanke en grondeienskappe, kan bestudeer en integreer. Ek het die impak van die teenwoordigheid van die indringer, E. camaldulensis in oewerbank areas langs die Bergrivier, wat ʼn Meditereense-tipe klimaat het, in die Wes Kaap, Suid Afrika, bestudeer. My studie het die volgende vrae gevra: 1) Is omgewings-omstandighede (bv. die oewerbank-omgewing geskep deur geomorfologiese prosesse) verander deur die teenwoordigheid van die Eucalyptus indringing, en indien wel, is hierdie omstandighede minder geskik vir inheemse spesies?; 2) Het die indringing van Eucalyptus gelei tot ʼn verandering in die samestelling van inheemse spesie gemeenskappe (moontlik beperkte plant voorplantings materiaal nadat die indringer verwyder is)?; 3) Hoe verander die indringing van Eucalyptus die samestelling van saad teenwoordig in die grond (saad bank) van die oewerbank en het dit die potensiaal om ekosisteem-herstel weer aan die gang te sit?; 4) Verander indringer Eucalyptus grond eienskappe (fisies en chemies)? Data vir al vier vrae is by dieselfde plek versamel om vergelykings te kan tref. Vir die eerste vraag het ek fotogrammetrie-tegnieke in ʼn Geografiese Inligting Stelsel (GIS) en afstandwaarnemings-omgewing toegepas om herhaalde lugfotos (1938-2010) te analiseer om die indringings geskiedenis sowel as die oewerbank se geomorfologiese evolusionêre dinamika saam te stel. Vir die tweede en derde vraag, het ek die aantal, diversiteit, gelykheid en samestelling van die bo-grond en saad bank plantegroei tussen natuurlike (geen indringers) en die gradiënt met indringers vergelyk. Vir die vierde en laaste vraag, het ek ʼn stel veranderlikes wat die fisies-chemiese eienskappe van die natuurlike omgewing sonder indringers beskryf, vergelyk met die eienskappe van die indringers gradiënt op ʼn seisoenale basis. Die resultate van die geomorfologiese studie wys dat die oewerbank omgewing baie dinamies was oor die afgelope 70 jaar met ʼn totale afname van 13.5% in rivieroewer-area in die gebied wat bestudeer is. Daar is ook bewyse vir ʼn vernouing van die kanaal, toename in die steilte van die rivierbank sowel as insnyding in die rivierbedding in areas wat ernstige indringing ervaar. Daar was geen beduidende neigings in die hidro-meteorologiese data nie. Die studie van die bo-grondse plantegroei het gewys dat die aantal plantspesies, diversiteit en strukturele kenmerke (bv. hoogte, dekking en gemiddelde basale diameter) van inheemse spesies afneem soos wat die indringings gradiënt toeneem. Indringing verander ook die inheemse sowel as uitheemse plant spesies samestelling. Beide inheemse en uitheemse spesies het areas wat slegs effense indringing wys, verkies. ʼn Studie van die saad bank (aantal sade in die grond) het gewys dat die indringer, E. camaldulensis geen merkwaardige invloed op die aantal, diversiteit en gelykheid van die totale en inheemse spesies teenwoordig gehad het nie. Die aantal uitheemse spesies was merkwaardig meer as die aantal inheemse spesies. Alhoewel indringing ook die samestelling van die inheemse saadbank beïnvloed het, was die inheemse saad bank, waarvan sommige spesies slegs daar voorgekom het, meer divers as die bo-grond plant samestellings. ʼn Studie van die grond het gewys dat die pH vlakke, afgesien van die seisoen, aansienlik laer is in die areas met indringers in vergeleke met areas sonder die indringers. Die hoeveelheid en dikte van plantmateriaal op die grond neem ook toe met ʼn toename in indringers. Die waterinhoud van die grond het afgeneem met ʼn toename in indringers en temperatuur het verhoog in beide winter en lente. Die totale makro- en mikrovoedingstowwe het nie merkwaardig gewissel saam met die indringingsgradiënt nie (p > 0.05), maar die uitruilbare katioon inhoud was aansienlik hoër, veral in die winter en lente-maande, in die areas sonder indringers in vergelyking met die areas met indringers. Vernouing van rivierkanaal en versteiling van rivieroewers verminder die geskiktheid van die oewersone om sekere ekologiese funksies te verrig soos grond-saadstoring en die verskaffing van ruimte vir plante om te groei. Dit verminder geleenthede vir werwing van saailinge en verminder gevolglik spesiediversiteit. Aan die ander kant, die vermindering van die spesiediversiteit in die bogrondse plantegroei verminder saadinsette in die grond en dit lei tot verarming van grondgestoorde saadbanke. Die oorvloed van sade van Acacia mearnsii in die grondgestoorde saadbank hou die bedreiging van sekondêre indringing in na die skoonmaak van E. camaldulensis indien dit wel as 'n bestuursaksie aangeneem word. Hierdie studie het die uiteengesette doelwitte suksesvol aangespreek. Die oorgang van ʼn oewerbank-gemeenskap wat hoofsaaklik uit inheemse plante bestaan na ʼn gemeenskap wat slegs uit E. camaldulensis bestaan, oor ʼn periode van ongeveer 50 jaar, het duidelike veranderinge tot gevolg gehad. Hierdie veranderinge is veral opmerklik in die oewerbank-geomorfologie en bo-grondse plantegroei, maar tot op hede is dit nog nie so opmerklik in die saadbank en grond-voedingstowwe konsentrasies nie. Hierdie bevindinge bied ondersteuning aan die WvW verwyderings-ondernemings en wys dat daar wel moontlike herstel van inheemse spesies kan wees na die verwydering van die indringer Eucalyptus. Hierdie kennis dra nie net by tot verhoogde effektiwiteit van herstelwerk en langtermyn-bestuursplanne nie, maar ook tot die navorsingsveld van oewerbankekologie en-herstel. Verdere eksperimentele navorsing word benodig om die moontlike meganismes wat verantwoordelik is vir die waargeneemde veranderinge, geassosieerd met Eucalyptus indringing, te verduidelik. Die bydrae van landbou tot die vorming van kanaal morfologie moet ook verder bestudeer word.