Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
A framework for modelling the spatio-temporal competition and spread of invasive plant species
Thesis (PhD)--Stellenbosch University, 2023.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | en_ZA en_ZA |
| Published: |
Stellenbosch : Stellenbosch University
2023
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613908367310848 |
|---|---|
| access_status_str | Open Access |
| author | Flemming, Alexander |
| author2 | Van Vuuren, Jan-Harm |
| author_browse | Flemming, Alexander Van Vuuren, Jan-Harm |
| author_facet | Van Vuuren, Jan-Harm Flemming, Alexander |
| author_sort | Flemming, Alexander |
| collection | Thesis |
| dc_rights_str_mv | Stellenbosch University |
| description | Thesis (PhD)--Stellenbosch University, 2023. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/126951 |
| institution | Stellenbosch University (South Africa) |
| language | en_ZA en_ZA |
| last_indexed | 2026-06-10T12:43:37.288Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2023 |
| publishDateRange | 2023 |
| publishDateSort | 2023 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/126951 A framework for modelling the spatio-temporal competition and spread of invasive plant species Flemming, Alexander Van Vuuren, Jan-Harm Stellenbosch University. Faculty of Engineering. Dept. of Industrial Engineering. Invasive plants -- South Africa Spatial data infrastructures Cellular automata Thesis (PhD)--Stellenbosch University, 2023. ENGLISH ABSTRACT: The study of introduced plant and animal species is important in areas which are rich in indigenous biodiversity. An introduced species is classified as a species living outside its native distribution range, but which has been introduced to the new environment either by accidental or deliberate human activity. Not all introduced species are able to survive in the new environment, but species that become established and spread beyond the place of introduction are reclassified as naturalised species. Not all naturalised species are harmful to the natural ecology or to humans. It is, therefore, important to distinguish between naturalised species that are either benign or useful, and those that become harmful and invasive. These classifications may indicate whether a species should be removed from its new environment and whether there are any environmental or ethical complications surrounding such a removal. Invasive alien species are a sub-category of naturalised species that negatively impact the natural species of an area. These invasions cause natural functioning ecosystems to break down, leading to further invasions and can ultimately lead to the extinction of the indigenous species of the area. The application of automated processes aimed at the construction of a generic framework capable of conducting four facets of invasive species modelling is pursued in this dissertation. The first is to conduct spatial analyses of ecosystems containing invasive species and to construct spatial data sets that map the distribution of these species in selected study regions. A machine learning algorithmic approach is followed in order to determine which environmental characteristics within a study region may be attributed to the distribution of invasive species within the ecosystem thereof. This approach is applied to predict areas that may require investigation due to likely unmapped occurrences or predicted future spread of invasive species, based on the area’s capability to sustain these species. Penultimately, a grid-based spatio-temporal modelling approach, known as cellular automata, is applied to simulate the competition between, and spread of, invasive species over a desired period of time. Finally, adaptive management strategies are applied with a view to attempt to inhibit the control of the spatio-temporal spread of invasive species infestations. The aforementioned automated processes combine the use of publicly available citizen science data pertaining to the distributions of invasive species, as well as data representing the environmental characterstics of countries and/or regions all over the world. Citizen science data collected on the iNaturalist website provides a large database of information on native as well as introduced plant and animal species globally, including type, location and time of identification. The data are collected by the public and verified by field experts in order to be suitable for use during academic research projects. Moreover, published research data of a high quality, including surveys of species collected in specific areas by field experts, and country-wide environmental characteristics (e.g land cover, soil, slope and elevation data) as well as climatic patterns are available in the literature. The framework proposed in this dissertation is validated by developing an instantiation of the framework in the form of a fully functional decision support system and applying this instantiation to a case study on the spatio-temporal competition between, and spread of, Australian Acacia species that are considered invasive in South Africa. The species-environment relationships extracted by the machine learning modelling approach are validated by comparing the model outputs with known ranges of habitat requirements of the Acacia species extracted from the literature. The cellular automata model results are further validated by visualising the population densities of the species at discrete time steps over a period of simulation. Finally, the resulting effects of recommended adaptive management strategies on the spatio-temporal spread of the invasive Australian Acacia species are visually compared with their uninhibited spread over the same time interval. AFRIKAANS OPSOMMING: Die studie van ingevoerde plant- en dierspesies is belangrik in areas wat ryk is aan inheemse biodiversiteit. ’n Ingevoerde spesie word geklassifiseer as ’n spesie wat buite sy inheemse verspreidingsgebied leef, maar wat ´of deur toevallige ´of doelbewuste menslike aktiwiteit in ’n nuwe omgewing bekendgestel is. Nie alle ingevoerde spesies is in staat om in die nuwe omgewing te oorleef nie, maar spesies wat daar vestig en buite die bekendstellingspunt versprei, word as genaturaliseerde spesies herklassifiseer. Nie alle genaturaliseerde spesies is skadelik vir die natuurlike ekologie of vir mense nie. Dit is dus belangrik om te onderskei tussen genaturaliseerde spesies wat goedaardig of nuttig is, en di´e wat skadelik en indringend word. Hierdie klassifikasies kan aandui of ’n spesie uit sy nuwe omgewing verwyder behoort te word en of daar enige omgewings- of etiese komplikasies rondom so ’n verwydering is. Indringer uitheemse spesies is ’n deelkategorie van genaturaliseerde spesies wat ’n negatiewe impak op die natuurlike spesies van ’n gebied het. Hierdie indringers mag veroorsaak dat natuurlik funksionerende ekosisteme in duie stort, mag lei tot verdere indringing en kan uiteindelik lei tot die uitsterwing van die inheemse spesies van die gebied. Die toepassing van ge-outomatiseerde prosesse wat gemik is op die konstruksie van ’n generiese raamwerk wat in staat is om vier fasette van indringerspesiemodellering te bewerkstellig, word in hierdie proefskrif nagestreef. Die eerste is om ruimtelike ontledings van ekosisteme uit te voer wat indringerspesies bevat en om ruimtelike datastelle te konstrueer wat die verspreiding van hierdie spesies in geselekteerde studiegebiede karteer. ’n Masjienleer-algoritmiese benadering word gevolg om te bepaal watter omgewingskenmerke binne ’n studiegebied aan die verspreiding van indringerspesies binne die ekosisteem toegeskryf kan word. Hierdie benadering word toegepas om gebiede te voorspel wat ondersoek mag vereis as gevolg van waarskynlike ongekarteerde voorkomste of voorspelde toekomstige verspreiding van indringerspesies, gebaseer op die gebied se vermo¨e om hierdie spesies te onderhou. Voorlaastens word ’n roostergebaseerde ruimteliktemporele modelleringsbenadering, bekend as sellulˆere outomate, toegepas om die mededinging tussen, en verspreiding van, indringerspesies oor ’n gespesifiseerde tydperk te simuleer. Uiteindelik word aanpasbare beheerstrategi¨e voorgestel en in ’n ruimtelik-temporele modelleringsbenadering toegepas om die verspreiding van die indringerspesies teen te werk. Die bovermelde ge-outomatiseerde prosesse kombineer die gebruik van publiek-beskikbare burger wetenskapdata wat betrekking het op die verspreidings van indringerspesies, asook data wat die omgewingskenmerke van lande en/of streke regoor die wˆereld verteenwoordig. Burger wetenskapdata wat op die iNaturalist-webwerf ingesamel is, verteenwoordig ’n groot databasis van inligting oor inheemse sowel as ingevoerde plant- en dierspesies wˆereldwyd, insluitend die tipe, ligging en tyd van identifikasie. Die data word deur die publiek ingesamel en deur spesialiste geverifieer om geskik te wees vir gebruik tydens akademiese navorsingsprojekte. Boonop is gepubliseerde navorsingsdata van ’n ho¨e gehalte in die literatuur beskikbaar, insluitend opnames van spesies wat in spesifieke gebiede deur spesialiste ingesamel is, asook landwye omgewingskenmerke (bv. grondbedekkings-, grond-, hellings- en hoogtedata) en klimaatspatrone. Die raamwerk wat in hierdie proefskrif voorgestel word, word deur ’n instansiasie daarvan in die vorm van ’n ten volle funksionele besluitsteunstelsel bekragtig en hierdie instansiasie word op ’n gevallestudie toegepas waarin die ruimtelik-temporele kompetisie tussen, en verspreiding van, Australiese Acacia spesies wat in Suid-Afrika as indringers beskou word. Die spesieomgewingsverwantskappe wat deur die masjienleer modelleringsbenadering onttrek word, word bekragtig deur die model uitsette met bekende intervalle van habitatsvereistes van die Acaciaspesies te vergelyk wat uit die literatuur verkry is. Die resultate van die sellulˆere outomatiese model word verder bekragtig deur die populasiedigthede van die pesies op diskrete tydstappe oor ’n simulasietydperk te visualiseer. Die finale effek van aanbevole aanpasbare beheerstrategi¨e op die ruimtelik-temporele verspreiding van die Australiese Acacia spesies word uiteindelik vergelyk met die onbeperkte verspreiding van die indringer spesies oor dieselfde simulasie tydperk. Doctorate 2023-02-06T07:58:24Z 2023-05-18T06:57:14Z 2023-02-06T07:58:24Z 2023-05-18T06:57:14Z 2023-02 Thesis http://hdl.handle.net/10019.1/126951 en_ZA en_ZA Stellenbosch University xxx, 321 pages : illustrations. application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Invasive plants -- South Africa Spatial data infrastructures Cellular automata Flemming, Alexander A framework for modelling the spatio-temporal competition and spread of invasive plant species |
| title | A framework for modelling the spatio-temporal competition and spread of invasive plant species |
| title_full | A framework for modelling the spatio-temporal competition and spread of invasive plant species |
| title_fullStr | A framework for modelling the spatio-temporal competition and spread of invasive plant species |
| title_full_unstemmed | A framework for modelling the spatio-temporal competition and spread of invasive plant species |
| title_short | A framework for modelling the spatio-temporal competition and spread of invasive plant species |
| title_sort | framework for modelling the spatio temporal competition and spread of invasive plant species |
| topic | Invasive plants -- South Africa Spatial data infrastructures Cellular automata |
| url | http://hdl.handle.net/10019.1/126951 |
| work_keys_str_mv | AT flemmingalexander aframeworkformodellingthespatiotemporalcompetitionandspreadofinvasiveplantspecies AT flemmingalexander frameworkformodellingthespatiotemporalcompetitionandspreadofinvasiveplantspecies |