{"id":{"repo_id":"stellenbosch","oai_identifier":"oai:scholar.sun.ac.za:10019.1/135624"},"canonical_url":"https://search.dev.ndltd.org/etd/stellenbosch/oai:scholar.sun.ac.za:10019.1/135624","repository":{"repo_id":"stellenbosch","name":"Stellenbosch University","base_url":"https://scholar.sun.ac.za/server/oai/request"},"display":{"title":"Development and testing of climate-sensitive site index models for long rotation Eucalyptus grandis in Limpopo province of South Africa","abstract":"ENGLISH ABSTRACT: Accurate prediction of height growth and site productivity in Eucalyptus grandis plantations has become increasingly important as the forestry sector faces rising climatic variability and long-term shifts in temperature and rainfall regimes. Although numerous growth and yield studies exist for short-rotation pulpwood systems, limited research has focused on long-rotation sawtimber production in South Africa, particularly with respect to climate-sensitive modelling. This study develops and evaluates dynamic, climate-responsive Site Index (SI) and dominant-height models for long-rotation E. grandis in Limpopo province, integrating long-term Permanent Sample Plot (PSP) data with high-resolution climatic datasets to support adaptive plantation management under changing environmental conditions. The study utilised PSP data collected by Merensky Timber across a broad range of altitudes, rainfall regimes, and productivity levels. Dominant height was calculated using the South African standard definition based on the mean height of the thickest 20% of trees. Climatic variables, including CHIRPS rainfall, CHIRTS temperature, potential evapotranspiration (PET), Standardised Precipitation-Evapotranspiration Index (SPEI), and the full suite of BIOCLIM indices, were spatially linked to each plot. Five dynamic height-growth models were fitted using non-linear least squares: Chapman–Richards (HT2CR4), Hossfeld (HT2HF3), Schumacher-type (HT2JC), McDill–Amateis (HT2MA2), and the log-Schumacher (HT2SCH) formulation. Model performance was assessed using separate calibration (70%) and validation (30%) PSP sets, followed by independent blind validation against compartment-level enumeration data. Parameter–climate relationships were quantified using correlation analysis and stepwise AIC model selection, providing the basis for developing climate-modified dynamic height equations. The PSP dataset captured wide variability in dominant height and site productivity, supporting robust model fitting. Across the five dynamic models, HT2MA2 and HT2HF3 achieved the highest predictive accuracy, characterised by low RMSE, minimal bias, and strong biological realism throughout the rotation. The Chapman–Richards and Schumacher-type models showed weaker fit and structural instability at older ages. Climate–parameter linkage analyses revealed strong associations between precipitation seasonality, dry- and wet-season rainfall (e.g., BIO16, BIO17), drought indices (SPEI), and key model parameters, indicating that climatic gradients significantly influence height-growth dynamics in long-rotation E. grandis. Despite clear climate–parameter relationships, the inclusion of additive climate modifiers did not improve model performance. Climate-modified predictions showed increased error and instability, particularly for HT2HF3, which exhibited sensitivity to parameter perturbation. Independent blind validation confirmed that unmodified models consistently outperformed modified versions when applied beyond the PSP calibration dataset. These results highlight the need for more flexible modelling approaches—such as hierarchical, hybrid, or process-informed frameworks—before climate variables can be reliably incorporated into operational height-growth projections. Overall, this study demonstrates that dynamic empirical models, especially HT2MA2 and HT2HF3, provide accurate and stable representations of dominant-height growth for long-rotation E. grandis in Limpopo. While climate clearly shapes height-growth potential, simple additive parameter modifiers are insufficient for operational deployment. The modelling framework developed here establishes a scientifically robust foundation for climate-aware growth forecasting and provides direction for future research into more sophisticated, adaptive modelling strategies.","abstract_html":"ENGLISH ABSTRACT: Accurate prediction of height growth and site productivity in Eucalyptus grandis plantations has become increasingly important as the forestry sector faces rising climatic variability and long-term shifts in temperature and rainfall regimes. Although numerous growth and yield studies exist for short-rotation pulpwood systems, limited research has focused on long-rotation sawtimber production in South Africa, particularly with respect to climate-sensitive modelling. This study develops and evaluates dynamic, climate-responsive Site Index (SI) and dominant-height models for long-rotation E. grandis in Limpopo province, integrating long-term Permanent Sample Plot (PSP) data with high-resolution climatic datasets to support adaptive plantation management under changing environmental conditions. The study utilised PSP data collected by Merensky Timber across a broad range of altitudes, rainfall regimes, and productivity levels. Dominant height was calculated using the South African standard definition based on the mean height of the thickest 20% of trees. Climatic variables, including CHIRPS rainfall, CHIRTS temperature, potential evapotranspiration (PET), Standardised Precipitation-Evapotranspiration Index (SPEI), and the full suite of BIOCLIM indices, were spatially linked to each plot. Five dynamic height-growth models were fitted using non-linear least squares: Chapman–Richards (HT2CR4), Hossfeld (HT2HF3), Schumacher-type (HT2JC), McDill–Amateis (HT2MA2), and the log-Schumacher (HT2SCH) formulation. Model performance was assessed using separate calibration (70%) and validation (30%) PSP sets, followed by independent blind validation against compartment-level enumeration data. Parameter–climate relationships were quantified using correlation analysis and stepwise AIC model selection, providing the basis for developing climate-modified dynamic height equations. The PSP dataset captured wide variability in dominant height and site productivity, supporting robust model fitting. Across the five dynamic models, HT2MA2 and HT2HF3 achieved the highest predictive accuracy, characterised by low RMSE, minimal bias, and strong biological realism throughout the rotation. The Chapman–Richards and Schumacher-type models showed weaker fit and structural instability at older ages. Climate–parameter linkage analyses revealed strong associations between precipitation seasonality, dry- and wet-season rainfall (e.g., BIO16, BIO17), drought indices (SPEI), and key model parameters, indicating that climatic gradients significantly influence height-growth dynamics in long-rotation E. grandis. Despite clear climate–parameter relationships, the inclusion of additive climate modifiers did not improve model performance. Climate-modified predictions showed increased error and instability, particularly for HT2HF3, which exhibited sensitivity to parameter perturbation. Independent blind validation confirmed that unmodified models consistently outperformed modified versions when applied beyond the PSP calibration dataset. These results highlight the need for more flexible modelling approaches—such as hierarchical, hybrid, or process-informed frameworks—before climate variables can be reliably incorporated into operational height-growth projections. Overall, this study demonstrates that dynamic empirical models, especially HT2MA2 and HT2HF3, provide accurate and stable representations of dominant-height growth for long-rotation E. grandis in Limpopo. While climate clearly shapes height-growth potential, simple additive parameter modifiers are insufficient for operational deployment. The modelling framework developed here establishes a scientifically robust foundation for climate-aware growth forecasting and provides direction for future research into more sophisticated, adaptive modelling strategies.","abstract_has_math":false,"creators":["Venter, Kuno"],"institution":"Stellenbosch : Stellenbosch University","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Drew, David Michael","Germishuizen, Ilaria"],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026-03","date_published":"2026-03","updated_at":"2026-07-24T04:40:09Z","subjects":[],"languages":["en"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://scholar.sun.ac.za/handle/10019.1/135624","outbound_label":"Repository record","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Drew, David Michael","Germishuizen, Ilaria"]},{"key":"dc:contributor.other","label":"Dc Contributor Other","values":["Stellenbosch University. Faculty of AgriSciences. Dept. of Forest and Wood Science."]},{"key":"dc:creator","label":"Author","values":["Venter, Kuno"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-04-02T10:58:57Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2026-04-02T10:58:57Z"]},{"key":"dc:date.issued","label":"Date","values":["2026-03"]},{"key":"dc:publisher","label":"Institution","values":["Stellenbosch : Stellenbosch University"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://scholar.sun.ac.za/handle/10019.1/135624"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Thesis (MScFor)--Stellenbosch University, 2026.","Venter, K. 2026. Development and testing of climate-sensitive site index models for long rotation Eucalyptus grandis in Limpopo province of South Africa. Unpublished masters thesis. Stellenbosch: Stellenbosch University [online]. Available: https://scholar.sun.ac.za/items/7975f1d2-ad3b-4c84-b29a-b0daf787151b"]},{"key":"dc:description.abstract","label":"Abstract","values":["ENGLISH ABSTRACT: Accurate prediction of height growth and site productivity in Eucalyptus grandis plantations has become increasingly important as the forestry sector faces rising climatic variability and long-term shifts in temperature and rainfall regimes. Although numerous growth and yield studies exist for short-rotation pulpwood systems, limited research has focused on long-rotation sawtimber production in South Africa, particularly with respect to climate-sensitive modelling. This study develops and evaluates dynamic, climate-responsive Site Index (SI) and dominant-height models for long-rotation E. grandis in Limpopo province, integrating long-term Permanent Sample Plot (PSP) data with high-resolution climatic datasets to support adaptive plantation management under changing environmental conditions. The study utilised PSP data collected by Merensky Timber across a broad range of altitudes, rainfall regimes, and productivity levels. Dominant height was calculated using the South African standard definition based on the mean height of the thickest 20% of trees. Climatic variables, including CHIRPS rainfall, CHIRTS temperature, potential evapotranspiration (PET), Standardised Precipitation-Evapotranspiration Index (SPEI), and the full suite of BIOCLIM indices, were spatially linked to each plot. Five dynamic height-growth models were fitted using non-linear least squares: Chapman–Richards (HT2CR4), Hossfeld (HT2HF3), Schumacher-type (HT2JC), McDill–Amateis (HT2MA2), and the log-Schumacher (HT2SCH) formulation. Model performance was assessed using separate calibration (70%) and validation (30%) PSP sets, followed by independent blind validation against compartment-level enumeration data. Parameter–climate relationships were quantified using correlation analysis and stepwise AIC model selection, providing the basis for developing climate-modified dynamic height equations. The PSP dataset captured wide variability in dominant height and site productivity, supporting robust model fitting. Across the five dynamic models, HT2MA2 and HT2HF3 achieved the highest predictive accuracy, characterised by low RMSE, minimal bias, and strong biological realism throughout the rotation. The Chapman–Richards and Schumacher-type models showed weaker fit and structural instability at older ages. Climate–parameter linkage analyses revealed strong associations between precipitation seasonality, dry- and wet-season rainfall (e.g., BIO16, BIO17), drought indices (SPEI), and key model parameters, indicating that climatic gradients significantly influence height-growth dynamics in long-rotation E. grandis. Despite clear climate–parameter relationships, the inclusion of additive climate modifiers did not improve model performance. Climate-modified predictions showed increased error and instability, particularly for HT2HF3, which exhibited sensitivity to parameter perturbation. Independent blind validation confirmed that unmodified models consistently outperformed modified versions when applied beyond the PSP calibration dataset. These results highlight the need for more flexible modelling approaches—such as hierarchical, hybrid, or process-informed frameworks—before climate variables can be reliably incorporated into operational height-growth projections. Overall, this study demonstrates that dynamic empirical models, especially HT2MA2 and HT2HF3, provide accurate and stable representations of dominant-height growth for long-rotation E. grandis in Limpopo. While climate clearly shapes height-growth potential, simple additive parameter modifiers are insufficient for operational deployment. The modelling framework developed here establishes a scientifically robust foundation for climate-aware growth forecasting and provides direction for future research into more sophisticated, adaptive modelling strategies.","AFRIKAANSE OPSOMMING: Akkurate voorspelling van hoogtegroei en terreinproduktiwiteit in Eucalyptus grandis-plantasies het toenemend belangrik geword namate die bosboubedryf toenemende klimaatsvariasie en langtermynverskuiwings in temperatuur- en reënvalpatrone ervaar. Alhoewel talle groei- en opbrengsstudies bestaan vir kort-rotasie pulphout stelsels, is daar beperkte navorsing oor lang-rotasie saaghoutproduksie in Suid-Afrika, veral met betrekking tot klimaat-sensitiewe modellering. Hierdie studie ontwikkel en evalueer dinamiese, klimaat-responsiewe terreinindeks (SI) en dominante-hoogte modelle vir lang-rotasie E. grandis in die Limpopo-provinsie, deur langtermyn Permanente Proefperseel (PSP)-data te integreer met hoëresolusie-klimaatdatastelle om aanpasbare plantasiebestuur onder veranderende omgewingstoestande te ondersteun. Die studie het PSP-data gebruik wat deur Merensky Timber versamel is oor ’n wye reeks hoogtes, reënvastreke en produktiwiteitsvlakke. Dominante hoogte is bereken volgens die Suid-Afrikaanse standaarddefinisie, gebaseer op die gemiddelde hoogte van die dikste 20% bome. Klimaatveranderlikes, insluitend CHIRPS-reënval, CHIRTS-temperatuur, potensiële evapotranspirasie (PET), die Gestandaardiseerde Reënval-Evapotranspirasie-indeks (SPEI), en die volledige reeks BIOCLIM-indekse, is ruimtelik gekoppel aan elke proefperseel. Vyf dinamiese hoogtegroeimodelle is gepas met behulp van nie-lineêre kleinste-kwadrate: Chapman–Richards (HT2CR4), Hossfeld (HT2HF3), Schumacher-tipe (HT2JC), McDill–Amateis (HT2MA2), en die log-Schumacher (HT2SCH)-formulering. Modelprestasie is beoordeel deur afsonderlike kalibrasie- (70%) en validasie- (30%) PSP-stelle, gevolg deur onafhanklike blinde validasie teen kompartementvlak-opmetingsdata. Parameter-klimaatverhoudings is gekwantifiseer deur korrelasie-analise en stapsgewyse AIC-modelseleksie, wat die basis gevorm het vir die ontwikkeling van klimaat-gewysigde dinamiese hoogtevergelykings. Die PSP-datastel het ’n wye variasie in dominante hoogte en terreinproduktiwiteit vasgevang, wat robuuste modelpassing ondersteun. Oor die vyf dinamiese modelle heen het HT2MA2 en HT2HF3 die hoogste voorspellingsakkuraatheid behaal, gekenmerk deur lae RMSE, minimale vooroordeel en sterk biologiese realisme deur die hele rotasie. Die Chapman–Richards- en Schumacher-tipe modelle het swakker passing en strukturele onstabiliteit op ouer ouderdomme getoon. Klimaat-parameter-koppelingsanalises het sterk assosiasies tussen reënvalseisoenaliteit, droë- en natseisoenreënval (bv. BIO16, BIO17), droogte-indekse (SPEI), en sleutelmodelparameters onthul, wat aandui dat klimaatsgradiënte ’n beduidende invloed het op hoogtegroeidinamika in lang-rotasie E. grandis. Ten spyte van duidelike klimaat-parameterverhoudings, het die insluiting van additiewe klimaatmodifiseerders nie modelprestasie verbeter nie. Klimaat-gewysigde voorspellings het verhoogde fout en onstabiliteit getoon, veral vir HT2HF3, wat sensitiwiteit vir parameterversteuring vertoon het. Onafhanklike blinde validasie het bevestig dat ongemodifiseerde modelle konsekwent beter presteer het as gewysigde weergawes wanneer dit buite die PSP-kalibrasiedatastel toegepas is. Hierdie resultate beklemtoon die behoefte aan meer buigsame modelleringsbenaderings—soos hiërargiese, hibriede of proses-geïnformeerde raamwerke—voordat klimaatveranderlikes betroubaar in operasionele hoogtegroeiprojektering geïntegreer kan word. Algeheel toon hierdie studie dat dinamiese empiriese modelle, veral HT2MA2 en HT2HF3, akkurate en stabiele voorstellings van dominante-hoogtegroei vir lang-rotasie E. grandis in Limpopo bied. Alhoewel klimaat duidelik die hoogtegroeipotensiaal vorm, is eenvoudige additiewe parametermodifiseerders onvoldoende vir operasionele implementering. Die modelleringsraamwerk wat hier ontwikkel is, vestig ’n wetenskaplik robuuste grondslag vir klimaat-bewuste groeivoorspelling en bied rigting vir toekomstige navorsing na meer gesofistikeerde, aanpasbare modelleringsstrategieë."]},{"key":"dc:title","label":"Title","values":["Development and testing of climate-sensitive site index models for long rotation Eucalyptus grandis in Limpopo province of South Africa"]}]}],"canonical_facts":{"dc:contributor.advisor":["Drew, David Michael","Germishuizen, Ilaria"],"dc:contributor.other":["Stellenbosch University. Faculty of AgriSciences. Dept. of Forest and Wood Science."],"dc:creator":["Venter, Kuno"],"dc:date.accessioned":["2026-04-02T10:58:57Z"],"dc:date.available":["2026-04-02T10:58:57Z"],"dc:date.issued":["2026-03"],"dc:description":["Thesis (MScFor)--Stellenbosch University, 2026.","Venter, K. 2026. Development and testing of climate-sensitive site index models for long rotation Eucalyptus grandis in Limpopo province of South Africa. Unpublished masters thesis. Stellenbosch: Stellenbosch University [online]. Available: https://scholar.sun.ac.za/items/7975f1d2-ad3b-4c84-b29a-b0daf787151b"],"dc:description.abstract":["ENGLISH ABSTRACT: Accurate prediction of height growth and site productivity in Eucalyptus grandis plantations has become increasingly important as the forestry sector faces rising climatic variability and long-term shifts in temperature and rainfall regimes. Although numerous growth and yield studies exist for short-rotation pulpwood systems, limited research has focused on long-rotation sawtimber production in South Africa, particularly with respect to climate-sensitive modelling. This study develops and evaluates dynamic, climate-responsive Site Index (SI) and dominant-height models for long-rotation E. grandis in Limpopo province, integrating long-term Permanent Sample Plot (PSP) data with high-resolution climatic datasets to support adaptive plantation management under changing environmental conditions. The study utilised PSP data collected by Merensky Timber across a broad range of altitudes, rainfall regimes, and productivity levels. Dominant height was calculated using the South African standard definition based on the mean height of the thickest 20% of trees. Climatic variables, including CHIRPS rainfall, CHIRTS temperature, potential evapotranspiration (PET), Standardised Precipitation-Evapotranspiration Index (SPEI), and the full suite of BIOCLIM indices, were spatially linked to each plot. Five dynamic height-growth models were fitted using non-linear least squares: Chapman–Richards (HT2CR4), Hossfeld (HT2HF3), Schumacher-type (HT2JC), McDill–Amateis (HT2MA2), and the log-Schumacher (HT2SCH) formulation. Model performance was assessed using separate calibration (70%) and validation (30%) PSP sets, followed by independent blind validation against compartment-level enumeration data. Parameter–climate relationships were quantified using correlation analysis and stepwise AIC model selection, providing the basis for developing climate-modified dynamic height equations. The PSP dataset captured wide variability in dominant height and site productivity, supporting robust model fitting. Across the five dynamic models, HT2MA2 and HT2HF3 achieved the highest predictive accuracy, characterised by low RMSE, minimal bias, and strong biological realism throughout the rotation. The Chapman–Richards and Schumacher-type models showed weaker fit and structural instability at older ages. Climate–parameter linkage analyses revealed strong associations between precipitation seasonality, dry- and wet-season rainfall (e.g., BIO16, BIO17), drought indices (SPEI), and key model parameters, indicating that climatic gradients significantly influence height-growth dynamics in long-rotation E. grandis. Despite clear climate–parameter relationships, the inclusion of additive climate modifiers did not improve model performance. Climate-modified predictions showed increased error and instability, particularly for HT2HF3, which exhibited sensitivity to parameter perturbation. Independent blind validation confirmed that unmodified models consistently outperformed modified versions when applied beyond the PSP calibration dataset. These results highlight the need for more flexible modelling approaches—such as hierarchical, hybrid, or process-informed frameworks—before climate variables can be reliably incorporated into operational height-growth projections. Overall, this study demonstrates that dynamic empirical models, especially HT2MA2 and HT2HF3, provide accurate and stable representations of dominant-height growth for long-rotation E. grandis in Limpopo. While climate clearly shapes height-growth potential, simple additive parameter modifiers are insufficient for operational deployment. The modelling framework developed here establishes a scientifically robust foundation for climate-aware growth forecasting and provides direction for future research into more sophisticated, adaptive modelling strategies.","AFRIKAANSE OPSOMMING: Akkurate voorspelling van hoogtegroei en terreinproduktiwiteit in Eucalyptus grandis-plantasies het toenemend belangrik geword namate die bosboubedryf toenemende klimaatsvariasie en langtermynverskuiwings in temperatuur- en reënvalpatrone ervaar. Alhoewel talle groei- en opbrengsstudies bestaan vir kort-rotasie pulphout stelsels, is daar beperkte navorsing oor lang-rotasie saaghoutproduksie in Suid-Afrika, veral met betrekking tot klimaat-sensitiewe modellering. Hierdie studie ontwikkel en evalueer dinamiese, klimaat-responsiewe terreinindeks (SI) en dominante-hoogte modelle vir lang-rotasie E. grandis in die Limpopo-provinsie, deur langtermyn Permanente Proefperseel (PSP)-data te integreer met hoëresolusie-klimaatdatastelle om aanpasbare plantasiebestuur onder veranderende omgewingstoestande te ondersteun. Die studie het PSP-data gebruik wat deur Merensky Timber versamel is oor ’n wye reeks hoogtes, reënvastreke en produktiwiteitsvlakke. Dominante hoogte is bereken volgens die Suid-Afrikaanse standaarddefinisie, gebaseer op die gemiddelde hoogte van die dikste 20% bome. Klimaatveranderlikes, insluitend CHIRPS-reënval, CHIRTS-temperatuur, potensiële evapotranspirasie (PET), die Gestandaardiseerde Reënval-Evapotranspirasie-indeks (SPEI), en die volledige reeks BIOCLIM-indekse, is ruimtelik gekoppel aan elke proefperseel. Vyf dinamiese hoogtegroeimodelle is gepas met behulp van nie-lineêre kleinste-kwadrate: Chapman–Richards (HT2CR4), Hossfeld (HT2HF3), Schumacher-tipe (HT2JC), McDill–Amateis (HT2MA2), en die log-Schumacher (HT2SCH)-formulering. Modelprestasie is beoordeel deur afsonderlike kalibrasie- (70%) en validasie- (30%) PSP-stelle, gevolg deur onafhanklike blinde validasie teen kompartementvlak-opmetingsdata. Parameter-klimaatverhoudings is gekwantifiseer deur korrelasie-analise en stapsgewyse AIC-modelseleksie, wat die basis gevorm het vir die ontwikkeling van klimaat-gewysigde dinamiese hoogtevergelykings. Die PSP-datastel het ’n wye variasie in dominante hoogte en terreinproduktiwiteit vasgevang, wat robuuste modelpassing ondersteun. Oor die vyf dinamiese modelle heen het HT2MA2 en HT2HF3 die hoogste voorspellingsakkuraatheid behaal, gekenmerk deur lae RMSE, minimale vooroordeel en sterk biologiese realisme deur die hele rotasie. Die Chapman–Richards- en Schumacher-tipe modelle het swakker passing en strukturele onstabiliteit op ouer ouderdomme getoon. Klimaat-parameter-koppelingsanalises het sterk assosiasies tussen reënvalseisoenaliteit, droë- en natseisoenreënval (bv. BIO16, BIO17), droogte-indekse (SPEI), en sleutelmodelparameters onthul, wat aandui dat klimaatsgradiënte ’n beduidende invloed het op hoogtegroeidinamika in lang-rotasie E. grandis. Ten spyte van duidelike klimaat-parameterverhoudings, het die insluiting van additiewe klimaatmodifiseerders nie modelprestasie verbeter nie. Klimaat-gewysigde voorspellings het verhoogde fout en onstabiliteit getoon, veral vir HT2HF3, wat sensitiwiteit vir parameterversteuring vertoon het. Onafhanklike blinde validasie het bevestig dat ongemodifiseerde modelle konsekwent beter presteer het as gewysigde weergawes wanneer dit buite die PSP-kalibrasiedatastel toegepas is. Hierdie resultate beklemtoon die behoefte aan meer buigsame modelleringsbenaderings—soos hiërargiese, hibriede of proses-geïnformeerde raamwerke—voordat klimaatveranderlikes betroubaar in operasionele hoogtegroeiprojektering geïntegreer kan word. Algeheel toon hierdie studie dat dinamiese empiriese modelle, veral HT2MA2 en HT2HF3, akkurate en stabiele voorstellings van dominante-hoogtegroei vir lang-rotasie E. grandis in Limpopo bied. Alhoewel klimaat duidelik die hoogtegroeipotensiaal vorm, is eenvoudige additiewe parametermodifiseerders onvoldoende vir operasionele implementering. Die modelleringsraamwerk wat hier ontwikkel is, vestig ’n wetenskaplik robuuste grondslag vir klimaat-bewuste groeivoorspelling en bied rigting vir toekomstige navorsing na meer gesofistikeerde, aanpasbare modelleringsstrategieë."],"dc:identifier.uri":["https://scholar.sun.ac.za/handle/10019.1/135624"],"dc:language.iso":["en"],"dc:publisher":["Stellenbosch : Stellenbosch University"],"dc:title":["Development and testing of climate-sensitive site index models for long rotation Eucalyptus grandis in Limpopo province of South Africa"],"dc:type":["Thesis"]},"updated_at":"2026-07-24T04:40:09Z"}