Science: Theses and Dissertations

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 20 of 5073
  • ItemOpen Access
    Neuromorphic computing with percolating networks of nanoparticles.
    (2025) Steel, Jamie
    Conventional computing technology, comprising strict regular arrays of transistors, has evolved into an integral component of modern society, significantly influencing many daily human activities. The demand for these devices continues to rise, alongside the complexity of the tasks they are designed to address. Recent developments have illuminated the limitations of conventional architectures, driving deep comparisons between traditional computers and the human brain. The brain’s exceptional computational efficiency has inspired the development of neuromorphic computing architectures, a paradigm in computing that draws inspiration from biological systems. Various approaches to neuromorphic computing exist, encompassing both software and hardware innovations. A novel method that emulates the complex connectivity and critical dynamics of the brain involves the use of percolating networks of nanoparticles (PNNs), which exhibit emergent characteristics analogous to those observed in biological neural networks. Hence, PNNs show promise for brain-like computation. This thesis examines the inherent complexity of the electrical signals measured from atomicscale switching activity in PNNs. Utilising several different measures, it is demonstrated that the internal dynamics of PNNs are complex and distinctly non-random. This complexity is present in the spatial, temporal, and spatio-temporal domains, and is consistently observed across multiple devices. Reservoir computing (RC) is a computing framework that leverages the temporal activity of a recurrent network to solve specific computational tasks. The rich network dynamics, obtained due to the correlated temporal signals measured at multiple network contacts, posit PNNs as a suitable physical reservoir. It is shown that PNNs utilised as reservoirs can successfully perform the benchmark task of spoken digit recognition with a classification accuracy comparable to the best-achieving physical neuromorphic systems. i
  • ItemOpen Access
    Zoospores-on-a-Chip : the development of a microfluidic platform to measure electrotactic responses of oomycete zoospores.
    (2024) Sarkar, Debolina
    Oomycetes are filamentous microorganisms that are morphologically and nutritionally similar to fungi. Some oomycete species, most notably those in the Phytophthora genus, can become pathogenic, and these can have huge impacts on ecosystems with the potential loss of biodiversity. Consequentially, this can have devastating impacts on primary industries such as agriculture, forestry, horticulture, viticulture, and aquaculture affecting food biosecurity. They can also have profound social and cultural effects. Because of their diversity, acquired resistance, and ease of spread, they can be extremely difficult to treat. In addition, there is still much to learn about their invasive mechanisms. Root-targeting members of the genus Phytophthora can spread via motile flagellated zoospores. These zoospores use a variety of mechanisms, including electrotaxis and chemotaxis, to locate plant root surfaces and undergo encystment. In previous studies, the electrotactic behaviour of bulk populations of zoospores has been characterised, but the molecular mechanisms by which electrotaxis takes place and the proteins involved in this process remain unknown. Chemotaxis is likely to involve the interception of root exudates by receptors in the zoospore membrane, although the exact mechanisms involved are also unknown. Furthermore, the relative contribution of these two processes to the overall tactic behaviour of zoospores is also unclear. Consequently, this study aimed to develop a microfluidic platform to increase understanding of biomolecular mechanisms underpinning electrotaxis and chemotaxis, processes that enable zoospores to locate plant roots. This involved the development of a microfluidic chip mimicking the plant root environment to observe zoospore behaviour under controlled electric fields. The three main objectives of this thesis were: 1) To design a chip with electrodes for observing zoospore swimming patterns and normal physiological processes (encystment/germination) under electric fields, 2) To use the chip to determine individual zoospore responses to varying electric field strengths and polarities, and 3) To develop a new chip to explore how electric fields and root-derived chemical attractants work together to influence zoospore movement. First, the change in dimension of the oomycete zoospores at different developmental stages was measured, which was crucial for studying their infective stages. Two resistive pulse sensing (TRPS) devices, qNano and qMicro, were used to measure the size and concentration of Phytophthora nicotianae zoospores. qNano initially revealed three particle populations, likely due to cell debris, motile zoospores, and cysts/germinating cysts. However, its limited size detection (unable to measure particles larger than 11 m) necessitated the use of qMicro for larger particles. Microscopic observations confirmed the presence of zoospores at various developmental stages, including cysts (average size - 15.25 μm), germinating cysts, and aggregated cysts. The use of qMicro enabled the successful measurement of motile zoospore and cyst sizes, with cysts being slightly larger (13.35 - 15.93 μm ± 0.94 μm) than motile zoospores (9.33 - 10.95 μm ± 0.55 μm). These findings were used to design the microfluidic platform that was used in this thesis work. A microfluidic platform for studying P. nicotianae zoospore electrotaxis was developed using photolithography and soft lithography. A silicon mould with desired channel patterns was created and used to cast a PDMS (polydimethylsiloxane) microfluidic layer. A combination of negative and positive photoresists was used for incorporating microstructures. An electrode substrate with a compatible layout was fabricated using separate photoresist techniques and wet etching. The PDMS layer and electrode substrate were then aligned and bonded for final assembly. The polarity of the electrodes and the generated field strength were controlled using a graphical user interface (GUI) and a customised set of electronics. Zoospores were introduced through designated inlets for observation within the micro-chamber under controlled electric fields generated by the electrodes. Zoospores were observed to be attracted to the electric field, spending more time in regions with stronger fields. The zoospores' movement patterns around the electrodes at different voltages were also tracked. The first-generation platform had some limitations. The electrodes were made of gold, which were too thin and easily damaged. The channel was also too long for the zoospores. The electronics was not efficient and did not always generate the expected electric field. The second-generation platform addressed these limitations. The electrodes were made thicker and connected in a loop structure to improve durability and connectivity. The electronics was also improved to allow for a wider range of voltages and more precise control of the electric field. With this new design, it was possible to observe that zoospores swam slower and exhibited a swirling motion near the cathode, and a repulsion zone was observed around the anode. The zoospores responded more quickly to the electric field at higher voltages. Next, a microfluidic platform was developed to examine how electric fields affect zoospores’ electrotactic responses in nature by arranging the electrodes in a root-tip-shaped structure. Unlike the second-generation device, the new platform featured five cathodes arranged around a root tip-like structure and one anode at the bottom, allowing for simultaneous application of five distinct electric fields. These electrodes had enhanced durability due to a thicker gold layer (100 nm), making them less prone to damage during the experiment. Their looped design allowed verification of each electrode pair functioning correctly by checking resistance and voltage differences. Additionally, the fabrication process was streamlined by using a positive photoresist instead of chrome etching, enabling faster and more precise removal of unwanted material. A new set of electronics with superior accuracy and user-friendliness was developed to control the platform. This platform was used to evaluate zoospores’ electrotactic responses by assigning five different voltages (0V, 0.5V, 1V, 1.5V, and 2V) to each of the electrodes in the root tip-structure. The findings indicated that zoospores were attracted to higher electric fields, with a maximum voltage of 2V producing a 9.3 mV/cm electric field. This electrotactic behaviour was observed consistently over time, showing a dose-dependent response. Notably, higher voltages led to cyst bursting after 120 minutes, indicating a potential threshold for electric field strength that zoospores could tolerate. The platform effectively mimicked the natural root environment, aiding in the detailed study of zoospore behaviour under varying electric conditions. In order to investigate both the electrotactic and chemotactic responses of zoospores, a PDMS layer with a micro-chamber, microchannels, and micropillars on a glass slide was created. The design incorporated microchannels for introducing chemoattractants, and electrodes were arranged within the micro-chamber in a root-tip-shaped structure. Chemical channels were 50 μm wide and 500 μm long, while zoospore inlets and outlets were 100 μm long and 30 μm high, with five 5 μm wide micro-outlets for chemical diffusion. Zoospore and chemical inlets were connected to syringe pumps to manage flow rates (10 μL/min for zoospores and 5 μL/min for dye), with flow control to prevent excessive diffusion. This setup had the potential to allow for real-time adjustment and observation of zoospore responses to combined electrical and chemical stimuli, effectively integrating electrotactic and chemotactic controls within a microfluidic environment. This platform was tested by introducing different coloured food-grade dyes through the inlets. Initially, dye diffusion through the micro-outlets proved to be a problem, but connecting separate syringe pumps to each inlet improved control, allowing the dye to fill the microchannels and remain for an average of 15.4 minutes before diffusing into the micro-chamber. Introducing a second dye demonstrated similar behaviour, remaining in the micro-chamber for 10 minutes before diffusing into the microchannels. Due to time constraints preventing combined electrotactic and chemotactic measurements with zoospores, the experiments provided a proof-of-concept for the platform's capability to study biological processes like zoospore behaviour in the future.
  • ItemOpen Access
    Cognitive impairment in Parkinson’s disease : investigating grey matter and magnetic susceptibility changes .
    (2024) Harris, Abbie
    Primarily characterised as a movement disorder, Parkinson’s disease (PD) is a progressive neurodegenerative disease, extending far beyond motor symptoms to significantly impact cognitive function. Mild cognitive impairment (MCI) is a well-defined transitional state on the continuum, ranging from normal cognition (PD-N) to dementia (PDD), and represents a risk factor for PDD conversion. The neural changes underlying cognitive decline in PD remain elusive, and the varied pattern of decline between patients represents a difficult barrier in understanding and managing this facet of PD. Although functional changes will be important, we can also presume that, neurodegeneration is a critical facet of cognitive decline in PD. Iron dysregulation is thought to contribute to neurodegeneration through mechanisms such as oxidative stress and neuroinflammation and magnetic resonance imaging (MRI) enables visualisation of in vivo pathological changes. Specifically, T1-weighted imaging (T1w) and quantitative susceptibility mapping (QSM), enable quantification of neuroanatomical changes and magnetic properties of brain tissue, respectively. We assessed the association among cognitive ability, grey matter (GM) volume (n=166, PD n=123; HC n=43), iron accumulation (n=96, PD n=70; HC n=26), and a hybrid image (n=89, PD n=62, HC n =27) based on T1w and QSM images to leverage the relative strengths of T1w and QSM images (TQ-SILiCON- T1w-QSM synthetic images via a linearly-weighted combination of contrasts). Cognition was assessed using detailed neuropsychological assessments including computerised and non-computerised cognitive batteries. We implemented three separate analysis approaches across the three imaging techniques: (1) a standard whole-brain general linear model approach, (3) source-based morphometry, a multivariate analysis approach, and (3) a region of interest analysis focusing on the hippocampus. We investigated T1w, QSM, and hybrid images for any differences (1) between HC and cognitive subgroups (HC, PD-N, PD-MCI, PDD), and (2) an association with cognitive ability (defined via standard neuropsychological assessment and via a computerised cognitive battery). GM atrophy increased monotonically from HC to PD-N, PD-MCI, and PDD, involving widespread cortical regions. These included the medial temporal lobe and limbic system structures such as the hippocampus, amygdala, and thalamus which showed a positive association with global cognition (FWE-corrected p < 0.05). Atrophy of hippocampal subfields, including the Cornu Ammonis regions and dentate gyrus, was identified between PD subgroups and was associated with global cognition in PD. In PD subgroups, we reported magnetic susceptibility alterations across the whole-brain, mainly located in subcortical GM regions such as the caudate, putamen, globus pallidus, substantia nigra, and red nucleus. ROI analysis of the hippocampus identified magnetic susceptibility changes across several subfields in PD subgroups which were negatively associated with global cognition in PD. We reported expected trends in computerised cognitive tasks including pattern separation, spatial and numeric WM tasks with progressive decreases from HC to PD-N to PD-MCI, although we did not find strong evidence for an association between these tasks and hippocampal integrity in PD. Fusing T1w and QSM images to reconstruct TQ-SILiCON images resulted in improved subcortical contrast of GM structures, specifically in the bilateral globus pallidus, substantia nigra pars compacta, substantia nigra pars reticula, and red nucleus, suggesting potential utility of the fusion method for future PD research. Increased GM atrophy and iron accumulation likely occur as cognitive severity increases in PD, but whether this is a cause or consequence of disease pathology remains unknown. Considering the well-established association between PD-MCI and PDD conversion, quantifying small changes in these brain metrics in PD-N and PD-MCI may further our understanding of neural changes underlying pathology.
  • ItemOpen Access
    Recurrent self-harm in New Zealand prisons : who is at greatest risk?
    (2024) Archbold, Matthew
    Self-harm within the prison system is an under researched area, especially in the context of New Zealand. Previous research focusing internationally has found that there are multiple risk factors for self-harm while incarcerated, these risk factors include social interactions in prison, mental health, and prior histories of self-harming behaviours. This study expanded the knowledge of self-harm behaviours in the New Zealand context. Offenders who were incarcerated between 2016 and 2020 and had at least one incident of self-harm were included in the sample. Factors such as age, gender, custody status, and duration of incarceration were included in a latent class analysis to identify any underlying groups that were present in the data. The latent class analysis identified three classes of individuals based on self-harm behaviours: two low-risk classes and one high-risk class. The high-risk class had an over-representation of individuals with offences against property, females, and individuals on remand. The key differences between the three classes was the number of incidents per prisoner and the average duration of incarceration for each group. Overall, findings showed that a minority of individuals had the greatest risk of self-harm (the high-risk class). In addition, the majority of incidents occurred in the earliest months of incarceration, suggesting that this is a high-risk period for prisoners and that there is some form of adjustment period occurring when an offender enters the prison system; this was supported by the majority of offenders having one or fewer prior imprisonments, suggesting that prison was a new environment for most prisoners in the sample. Future research is required for greater understanding of the relationship between these two factors, particularly in the New Zealand context. What this study found largely aligns with previous research that a minority of individuals account for the majority of self-harm incidents, and that there is an overrepresentation of females in the self-harming data.
  • ItemOpen Access
    Entering out-of-home care in Aotearoa, New Zealand: a qualitative inquiry into the recalled experiences of young people
    (2024) Cagney, Ellen
    BACKGROUND: For children who have been maltreated, being removed from their family of origin constitutes the highest level of government intrusiveness and intervention. As of June 2023, there were a total of 4,317 children and young people in out-of-home care (OOHC) in Aotearoa, New Zealand. The subjective experiences of children and young people entering OOHC have not been formally researched in Aotearoa, New Zealand. AIM: Therefore, the present study aimed to explore the recalled experiences of entering OOHC as a child or young person in Aotearoa, New Zealand. METHODOLOGY: Through a phenomenological inquiry, ten in-depth semi-structured interviews were conducted with care-experienced adults aged 18 to 46 years, and data were analysed thematically. FINDINGS: Analysis revealed five themes: (1) the familiar versus the unfamiliar, (2) the impact of connection, (3) the benefits of knowledge, (4) the power of communication and (5) the role of individuality. CONCLUSION: Participants emphasized the need for caregivers to show compassion and understanding for the past and current challenges of the children who are living with them. Findings highlight how much children entering OOHC do not understand about what is happening to them, and how distressing this is. Therefore, it is important children are kept informed and treated with kindness.
  • ItemOpen Access
    Management of birds on Canterbury farms : hedgerow use and deterrent efficacy
    (2024) Lowen, Fergus
    A large portion of the global land area is dedicated to agriculture, and it has implications for biodiversity. Restoration of habitat is an important step in maintaining biodiversity. Farms are now planting habitat that supports biodiversity. The first experimental component of my study investigated how hedgerow composition (native or exotic) influenced bird populations. I completed surveys of 13 hedgerow pairs (native and exotic) across the Canterbury Plains in the summer of 2022/2023 and the autumn of 2023. I found that more native birds were present in native hedgerows in autumn but there was no difference between hedgerows in the summer and temperature was the primary driver of bird abundance. Not all birds are desirable on farms and the use of deterrent devices is a common practice. However, the effectiveness of these deterrents has not been compared in an on-farm context in New Zealand. In the second experimental component of my study, I explain the methods I developed to make direct comparisons of common bird deterrent devices (gas gun, kite, and laser) against each other and a control. I used trail cameras to measure the number of birds and used this, as well as seed weight, to determine their effectiveness. I also compared seed preferences under the difference deterrent pressures. The seeds I compared were barley, oilseed rape, radish, and sunflower. I found that the deterrents were generally effective, but this was partially dependent on the response variable I was measuring. I also found that sunflower was the most preferred seed type and barley was the least preferred seed type. Oilseed rape and radish had similar levels of interest in them. This study will also allow farmers to make more informed decisions on the deterrents they use as well as understanding how landscape factors and hedgerow compositions influence bird populations.
  • ItemOpen Access
    Suicidal behaviours across the lifespan from early adolescence to mid-adulthood in a New Zealand based cohort
    (2024) Rippin, Louise
    Suicidal behaviours are complex and multifaceted, resulting from a wide range of interacting genetic, psychological, psychiatric, social, economic, and cultural risk factors. The aim of this current study was to explore trajectories of individuals experiencing suicidal ideation and/or suicide attempts, compared to those who do not experience these behaviours. Furthermore, this study aimed to provide up-to-date knowledge of both the predictors and outcomes of suicide behaviours, drawing on data from the Christchurch Health and Development Study. Participants consisted of a New Zealand birth cohort of 1265 individuals, with measures obtained at multiple timepoints up to 40 years of age. Across the lifespan, the prevalence of suicidal ideation ranged from 1.8-9.5%, and suicide plans/attempts ranged from 2.2%-3%. Six childhood factors predicted latent trajectory classes of suicidal behaviours: parental care, peer attitudes, sexual abuse, novelty seeking, parental illegal drug use, and ethnicity. Latent trajectory membership strongly predicted other mental health outcomes in adulthood: depression, anxiety, nicotine-use disorder, alcohol-use disorder, and shorter intimate/significant relationships. This research adds to the current body of literature on suicidal behaviours, allowing for greater understanding of the factors that may enhance an individual’s risk of engagement in these behaviours.
  • ItemOpen Access
    Approaches to Photo-activated Cytotoxins
    (2024) Munasinghe, Rajika Madushanee
    In this research, our synthetic targets are Zn(II) porphyrin – Co(III) heterodinuclear systems, which can be explored as photoactivated cytotoxins in chemotherapy. Initially, in Chapter 2, we have synthesised the main Zn(II) porphyrin based electron donor components [Zn(tpp-NH2)] 2.05 and [Zn{(tpp-NH2)4}] 2.08 while optimising literature-reported methods. We have optimised methods to obtain the best yield while reducing the chemical cost and solvent usage. Specifically, [Zn(tpp-NH2)] 2.05 was of interest to us to elaborate and use as a donor component in the targeted Zn(II)-Co(III) heterodinuclear complexes. The majority of work focused on designing and synthesising elaborated Zn(II) porphyrin electron donor components. Chapter 3 describes the development and use of reductive amination reactions to elaborate [Zn(tpp-NH2)] 2.05. From this study, we have been able to synthesise four new zinc porphyrin complexes ([Zn(L1)], [Zn(L2)], [Zn(L3)], [Zn(L4)]) with the different bidentate or tridentate second metal binding site. Chapter 4 describes the construction of a second metal binding site on the amine group of [Zn(tpp-NH2)](2.05) using amide coupling reactions. From the Chapter 4 study, we have been able to synthesise one new zinc porphyrin complex ([Zn(L7)]) and two more literature-reported zinc porphyrins ([Zn(L6)] and [Zn (L8)]) while optimising literature methods. Chapter 5 explores the Co(III) complexation ability of those elaborated Zn(II) porphyrin-based electron donor components using cobalt(III) triflate (en, tren and dien) complexes. In this Chapter, we provide evidence for the formation of Zn(II) porphyrin – Co(III) heterodinuclear complexes by reacting [Zn(L1)] and [Co(en)2(OTf)2]OTf and then the photoactivated ligand release upon light exposure. Moreover, the formation of a heterodinuclear complex between [Zn(L8)] and [Co(dien)(OTf)3] was also promising based on the NMR spectroscopy and mass spectrometry evidence. In Chapter 6, we attempted to synthesise cytotoxic Co(III) nitrogen mustard complexes with labile ligands such as triflate (OTf) and nitrile (NCCH3). Specifically, in this Chapter, we have experimented with the synthesis of [Co(ceen)2(OTf)2]OTf (6.04) and [Co(ceen)2(NCCH3)2]OTf3 (6.05). Based on the mass spectrometry, we have synthesised [Co(ceen)2(OTf)2]OTf (6.04), but unfortunately, we could not isolate it in a pure form. Then we explored the nitrilium chemistry in order to synthesise alternative [Co(ceen)2(NCCH3)2]OTf3 (6.05) complex. Even though we have not yet been able to synthesise the targeted [Co(ceen)2(NCCH3)2]OTf3 (6.05) complex, we were able to isolate new complex, [Co(ceen)2(NCCH3)(NO2)]OTf2 (6.14). Based on the results, the use of nitrilium chemistry is promising for the synthesis of the targeted [Co(ceen)2(NCCH3)2]OTf3 (6.05) complex, but further optimisations are required Our study provides a synthetic strategy that can be used to synthesise and study a diverse range of Zn(II) porphyrin – Co(III) heterodinuclear systems. Moreover, the preparation of two upcoming publications is ongoing– one addressing the photoactivated ligand release in a Zn(II) porphyrin – Co(III) heterodinuclear system and the other involving new Co(III) nitrogen mustard-nitrile complexes.
  • ItemOpen Access
    Changes in the Rakaia River mouth barrier dynamics.
    (2024) Brennan, A.; Drummond, L.; Lu, L.; Ninnes, J.; Smith, S.
    This study investigates the evolution of the Rakaia River mouth barrier, the processes influencing its morphology and position, and its impact on the local Rakaia Huts community. This is done by analyzing 30 years of beach profile, river cross-section, and marine/fluvial data. Analysis of LiDAR using Arc GIS Pro along with data manipulation and presentation in Excel was used to enhance our understanding of barrier dynamics. Additionally, statistical analysis of time series data using R Studio and Excel allowed us to examine the influence of various factors such as river flow, wave action, sedimentation, and storm events, on the barrier's erosion and sediment transport patterns. Our findings revealed a significant shift in the barrier's positioning over time. Between 1991 and 2008, the barrier exhibited progradation of 134 m, advancing towards the sea. However, since 2008, it has undergone a notable retreat of 60 m. Meanwhile, the beaches on either side of the barrier have continued to prograde, highlighting the complexity of coastal processes and likely interplay of sediment transport from the river to the coast. These findings have significant implications for flood management practices and community planning in the Rakaia Huts region. By understanding the historical and ongoing changes in the barrier, local authorities can make informed decisions to mitigate potential risks and safeguard the community. This research contributes to our broader knowledge of coastal barrier systems, particularly the complex interaction of hydrological and coastal processes under both natural and human-induced influences. While this study does provide valuable insight, it is essential to acknowledge limitations. The reliance of the study on pre-collected data may have restricted the accuracy of the analysis. Future research could benefit from real-time data collection, or at a more sub-annual temporal scale, and the use of advanced techniques like drone-based aerial imaging and repeat LiDAR to ensure the data is accurate. In conclusion, this study identifies the timing and drivers of key shifts in barrier position for the Rakaia River mouth barrier. Increased knowledge and a thorough analysis of its evolution contributes to the current understanding of the hydrological and coastal processes shaping the barrier, and implications for the local community.
  • ItemOpen Access
    Assessing the impacts on agriculture from volcanic ashfall of the 2022 Hunga Tonga-Hunga Ha’apai eruption, Kingdom of Tonga.
    (2024) Auapaau, Fiapaipai Ruth
    Explosive volcanic eruptions can produce a range of hazards including ashfall, which can impact agriculture systems, threatening the economies and food security of subsistence communities. Previous case studies of volcanic ashfall impacts on agriculture have been recorded for non-tropical, mostly temperate contexts. Agricultural practices and systems in these previous studies are largely export focused with a mixture of horticultural, pastoral and cereal cultivation. In Tongatapu, the study area for this thesis, agricultural systems are characterised by subsistence mixed cropping, primarily for household and domestic consumption. Despite understanding disaster risk being a priority of the Sendai framework for Disaster Risk Management (A global accord for DRM), there is a distinct lack of research on the short- to long-term vulnerability of Pacific tropical agriculture to volcanic ashfall. This gap hinders the development of effective strategies to mitigate volcanic impacts on Pacific agricultural systems and the ability to adapt agricultural practices in Pacific countries. This study aims to address this knowledge gap using post-event volcanic hazard and impact data from the 2022 Hung Tonga-Hunga Ha’apai eruption event to develop and apply a framework for assessing ash impacts on Pacific agriculture. The framework used includes an analysis of field interview data (collected outside the scope of the present study) and a remote, post-event crop damage assessment of Tongatapu agriculture. Asset (i.e. crops), hazard (i.e. volcanic ashfall thickness) and impact (i.e damage state) attributes were defined for a Tongatapu agricultural context (e.g. types of crops grown, subsistence (i.e. kitchen garden or land parcel), intercropping/mixed cropping) and used to classify observations from post-event imagery. Results from the remote, post-event crop damage assessment show the highest damage state (DS3) mostly recorded for tree crops (n=201), including bananas and breadfruit, which is consistent with field interview data. Crop-specific variations were observed, including, crops such as vanilla and kava exhibiting a consistent damage state (all DS1) whilst others such as bananas and breadfruit have a varied damage state distribution (DS1-DS3). Under non-mixed cropping conditions, tree and root crops showed increases in moderate (DS2) to severe damage (DS3) levels (i.e. root crop in mixed DS2 at 68.8%, in non-mixed DS2 at 72.2%). Damage distribution based on phenology found 56.8% of observations were at the fruit or crop production stages during the ashfall event. Crops at vegetative and flowering stages had the highest frequency attributed to a single damage state – DS2 (57.1% and 68.7% respectively). For example, citrus crops were observed as flowering at the time of the event and sustained considerable losses because of this (a vulnerable stage in growth cycle). Other crops such as coffee were recorded in the analysis of field interview data as having severe losses, however, no coffee-based observations were available in the remote, post-event damage assessment to validate this. The key findings of this thesis are: 1) there is considerable value in using remote impact assessment approaches that supplement or validate field studies to track volcanic ashfall impacts impact on agriculture systems; and 2) the importance of mixed cropping/intercropping as a mitigation/adaptation strategy for improving agricultural resilience in Pacific Island nations exposed to volcanic ashfall. The findings from this research contribute to developing an understanding of the vulnerability of Pacific agriculture to volcanic ashfall, particularly crops important for local community sustenance. The present study represents a first attempt to utilise post-event imagery captured for non-impact assessment-specific purposes to assess damage to crops in a tropical context and provides a framework for future applications. However, this thesis is not without its limitations, which included the spatial bias towards semi-urban locations in Tongatapu and the scope which was restricted to Tongatapu despite there being reports of agricultural impacts in other islands. Therefore, future research should consider the applicability of the results in this study to the remote or outer islands that were impacted, thus reducing the uncertainty of these findings. Impacts results of this study can inform DRM strategies in Tonga, and other tropical Pacific Island nations exposed to volcanic ashfall hazard, to ultimately increase agricultural system resilience.
  • ItemOpen Access
    Understanding how ethical leadership & perceived organizational support effect employee wellbeing with recruiters.
    (2025) Acharya, Ketki
    This study investigates the relationship between workload and wellbeing among recruiters working in recruitment agencies across New Zealand, with a focus on the moderating effects of ethical leadership and perceived organizational support. Using a cross-sectional survey design, data was collected from recruiters across various recruitment agencies to assess whether ethical leadership and perceived organizational support could buffer the impact of workload on recruiters' wellbeing. Moderated regression analyses were conducted to explore these potential buffering effects. Contrary to expectations, the findings showed no significant moderating effects of ethical leadership or perceived organizational support on the relationship between workload and well-being. However, both ethical leadership and perceived organizational support demonstrated stronger direct effects on this relationship. These findings suggest that, for recruiters in New Zealand, the influence of ethical leadership and perceived organizational support were more important to wellbeing than workload.
  • ItemOpen Access
    Te mōhio ki te arotake tūraru, he tirohanga o te ao Māori = Understand risk assessment, a Māori worldview perspective.
    (2024) Rose, Jack
    For generations, Māori (an Indigenous people of Aotearoa – New Zealand) have described the Tūraru (risk), Pōkaikaha (uncertainty), and Putanga (consequence) associated with a range of events throughout Te Moana nui a Kiwa (the Pacific Ocean) and Aotearoa. Although Māori have lived with Tūraru for centuries, there is a significant gap in the integration of Kaupapa Māori (Māori approaches) and Te Ao Māori (Māori worldviews) into Arotake Tūraru (risk assessments). This Rangahau (research) provides insights to help address this gap by (1) analysing Te Ao Māori concepts of and approaches to Tūraru, then (2) applying these concepts to a Māori-centric Arotake Tūraru (risk assessment), and discussing the differences and similarities between this approach and contemporary approaches. It demonstrates how Kaupapa Māori, Te Ao Māori values, and Mātauranga Māori (Māori knowledge systems) can be incorporated into risk assessment processes alongside contemporary frameworks to provide insights into Māori-centric risks. The application of Te Ao Māori to a risk assessment suggests that by integrating Indigenous approaches into the study of risk and practice of risk assessment, Indigenous knowledge systems, with our understanding of local ecosystems and long-standing environmental stewardship practices, can offer valuable contributions to addressing and understanding climate change risk. It also suggests that to equitably and effectively address Ngā Tūraru Huringa Āhuarangi (climate change risks) in Aotearoa, we must incorporate diverse perspectives to ensure that risk assessments, subsequent climate policies and management strategies are culturally relevant and consider Indigenous knowledge systems and values.
  • ItemOpen Access
    Do people perceive phantom costs when attractive people show romantic interest in unattractive people?
    (2024) Loginova, Victoria
    Phantom costs are perceptions of hidden costs to an offer which seems too good to be true. Prior research on phantom costs has focused on an economic perspective, however, no prior research has studied phantom costs in the area of romantic attraction. The present paper addresses this gap in the literature. Study 1 shows the existence of phantom costs using a behavioural design, whereas the second study uses a vignette design to re-establish the presence of phantom costs, as well as shows that when given a sufficient explanation, phantom costs are mitigated.
  • ItemOpen Access
    Untangling processes and mechanisms of stability in complex landscapes : lessons and applications for braided rivers.
    (2024) Harris, Holly
    Knowledge of ecological stability is important for both advancing understanding of ecological systems and applying appropriate management levers. However, stability is difficult to investigate in dynamic, heterogenous landscapes, which are typically the most anthropogenically threatened ecosystems. Braided rivers are heterogeneous and dynamic but provide an ideal opportunity for testing the role of cross-scale stability mechanisms that will be integral for the conservation of these globally rare, but culturally important landscapes. First, I investigated how distinct habitats within a braided river landscape were connected through movement and feeding strategies of predators, and I used stable isotope analyses to examine their dietary patterns across isotopically distinct river habitats. Across birds, fish, and spiders, I identified multiple food-web structures that are known to weaken interaction strengths, aiding stability. These results indicated stabilising food-web structures primarily relied on the flood-induced landscape heterogeneity generating spatial variability in resources. Next, to identify underlying drivers of stability at local (habitat) and regional (whole river) scales, I investigated compositional and aggregate variability in terrestrial and aquatic invertebrates, and variability and synchrony between aquatic invertebrates and fish. Synchrony between local aquatic invertebrate communities was low, likely driven by local differences in flood disturbance. However, terrestrial invertebrates had high composition and biomass synchrony, consistent with more homogeneous terrestrial conditions (e.g. temperature). Fish biomass had lower variability and synchrony than aquatic invertebrates, indicating mobile predators likely benefit from spatially heterogeneous conditions. Preserving spatial environmental heterogeneity will therefore be important for maintaining landscape-scale stability across multiple trophic levels. Finally, I explored the contributions of disturbance-driven heterogeneity and adaptive foraging to stability. I described growth of resources, consumers and predators using a mechanistic model and mimicked a mobile predator with an adaptive predator foraging function. Pulsed disturbances varying in magnitude and spatial extent, were applied to this model. System stability was highly contextual and regional connectivity through adaptive predator foraging on its own was necessarily stabilising, but the combination of disturbance and adaptive foraging was. Again, this suggests that understanding spatial landscape contexts and synergies between predator movement, landscape heterogeneity, and disturbance will be important for predicting system responses to changing disturbance regimes. Overall, I have shown that complex landscapes have multiple mechanisms for resilience, and preserving heterogeneity in braided rivers is likely critical for maintaining regional stability of their species assemblages. Additionally, heterogeneity is likely important for reducing the risk of disturbances cascading between trophic levels. By identifying stability-conferring mechanisms in a rare and complex braided river landscape, my work increases knowledge of both a poorly understood landscape and complex ecological systems more broadly. Bridging the gap between theoretical ecology and practical applications will be important to successful management outcomes for complex landscapes like braided river, particularly in a world where ecosystems face increasing pressures from multiple anthropogenic drivers.not
  • ItemOpen Access
    Modelling earthquake-induced landslide impacts on infrastructure systems in Wellington.
    (2024) Harvey, Julia
    The proximity of the Greater Wellington Region (GWR) to numerous active faults poses a sizeable earthquake hazard that has implications for the whole of Aotearoa New Zealand. The 2022 National Seismic Hazard Model suggests multiple plausible, major earthquake scenarios that would have significant impacts for the region. Whilst seismic hazard in the GWR is relatively well-studied, the cascading hazards, particularly landslides, are less well understood. Despite the known impacts of earthquake-induced landslides (EQIL), and the acknowledgement of EQIL risk in existing seismic hazard analyses, the potential consequences of this cascading hazard have not been quantified at a high resolution in the GWR. This is especially important in the context of the 2015 Sendai Framework, where the need for multihazard approaches to disaster risk reduction has been emphasised. Furthermore, with the range of plausible earthquake scenarios in the GWR, traditional approaches to seismic hazard analysis are becoming progressively unsuitable to assess seismic impacts across multiple seismic events. Most seismic hazard and risk assessments for GWR to date have focussed on probabilistic estimates of hazard or a limited number of potential scenarios involving rupture of the Wellington Fault or the Hikurangi Subduction Zone. However, given the number of plausible earthquake scenarios that could affect GWR, understanding whether each scenario produces bespoke impacts or if consistent impacts are seen in multiple scenarios is critical for planning. Ensemble modelling utilises a range of plausible earthquake scenarios to assess the variability of seismic impacts for a given area, combining strengths of existing deterministic and probabilistic approaches and therefore presents a potential opportunity for advancing a holistic understanding of seismic risk in this region. This study assesses the variability in EQIL hazard and resulting impacts in the GWR for a plausible earthquake scenario ensemble. It uses a statistical landslide model by applying fuzzy logic within GIS to determine landslide hazard for ten plausible earthquake scenarios across the GWR. The exposure of the road network to landsliding across the earthquake scenario ensemble is then estimated, to assess potential disruption to key routes and emergency water access. Importantly, this research explores how this exposure and disruption might vary between earthquake scenarios. The results show the criticality of multi-hazard, multi-scenario approaches for assessing seismic risk; although the extent and severity of EQIL hazard potential was variable across the ensemble, plausible EQIL impacts were present within the GWR across every modelled scenario. Numerous low-specificity impacts have been identified The proximity of the Greater Wellington Region (GWR) to numerous active faults poses a sizeable earthquake hazard that has implications for the whole of Aotearoa New Zealand. The 2022 National Seismic Hazard Model suggests multiple plausible, major earthquake scenarios that would have significant impacts for the region. Whilst seismic hazard in the GWR is relatively well-studied, the cascading hazards, particularly landslides, are less well understood. Despite the known impacts of earthquake-induced landslides (EQIL), and the acknowledgement of EQIL risk in existing seismic hazard analyses, the potential consequences of this cascading hazard have not been quantified at a high resolution in the GWR. This is especially important in the context of the 2015 Sendai Framework, where the need for multihazard approaches to disaster risk reduction has been emphasised. Furthermore, with the range of plausible earthquake scenarios in the GWR, traditional approaches to seismic hazard analysis are becoming progressively unsuitable to assess seismic impacts across multiple seismic events. Most seismic hazard and risk assessments for GWR to date have focussed on probabilistic estimates of hazard or a limited number of potential scenarios involving rupture of the Wellington Fault or the Hikurangi Subduction Zone. However, given the number of plausible earthquake scenarios that could affect GWR, understanding whether each scenario produces bespoke impacts or if consistent impacts are seen in multiple scenarios is critical for planning. Ensemble modelling utilises a range of plausible earthquake scenarios to assess the variability of seismic impacts for a given area, combining strengths of existing deterministic and probabilistic approaches and therefore presents a potential opportunity for advancing a holistic understanding of seismic risk in this region. This study assesses the variability in EQIL hazard and resulting impacts in the GWR for a plausible earthquake scenario ensemble. It uses a statistical landslide model by applying fuzzy logic within GIS to determine landslide hazard for ten plausible earthquake scenarios across the GWR. The exposure of the road network to landsliding across the earthquake scenario ensemble is then estimated, to assess potential disruption to key routes and emergency water access. Importantly, this research explores how this exposure and disruption might vary between earthquake scenarios. The results show the criticality of multi-hazard, multi-scenario approaches for assessing seismic risk; although the extent and severity of EQIL hazard potential was variable across the ensemble, plausible EQIL impacts were present within the GWR across every modelled scenario. Numerous low-specificity impacts have been identifiedThe proximity of the Greater Wellington Region (GWR) to numerous active faults poses a sizeable earthquake hazard that has implications for the whole of Aotearoa New Zealand. The 2022 National Seismic Hazard Model suggests multiple plausible, major earthquake scenarios that would have significant impacts for the region. Whilst seismic hazard in the GWR is relatively well-studied, the cascading hazards, particularly landslides, are less well understood. Despite the known impacts of earthquake-induced landslides (EQIL), and the acknowledgement of EQIL risk in existing seismic hazard analyses, the potential consequences of this cascading hazard have not been quantified at a high resolution in the GWR. This is especially important in the context of the 2015 Sendai Framework, where the need for multihazard approaches to disaster risk reduction has been emphasised. Furthermore, with the range of plausible earthquake scenarios in the GWR, traditional approaches to seismic hazard analysis are becoming progressively unsuitable to assess seismic impacts across multiple seismic events. Most seismic hazard and risk assessments for GWR to date have focussed on probabilistic estimates of hazard or a limited number of potential scenarios involving rupture of the Wellington Fault or the Hikurangi Subduction Zone. However, given the number of plausible earthquake scenarios that could affect GWR, understanding whether each scenario produces bespoke impacts or if consistent impacts are seen in multiple scenarios is critical for planning. Ensemble modelling utilises a range of plausible earthquake scenarios to assess the variability of seismic impacts for a given area, combining strengths of existing deterministic and probabilistic approaches and therefore presents a potential opportunity for advancing a holistic understanding of seismic risk in this region. This study assesses the variability in EQIL hazard and resulting impacts in the GWR for a plausible earthquake scenario ensemble. It uses a statistical landslide model by applying fuzzy logic within GIS to determine landslide hazard for ten plausible earthquake scenarios across the GWR. The exposure of the road network to landsliding across the earthquake scenario ensemble is then estimated, to assess potential disruption to key routes and emergency water access. Importantly, this research explores how this exposure and disruption might vary between earthquake scenarios. The results show the criticality of multi-hazard, multi-scenario approaches for assessing seismic risk; although the extent and severity of EQIL hazard potential was variable across the ensemble, plausible EQIL impacts were present within the GWR across every modelled scenario. Numerous low-specificity impacts have been identified across the GWR; as a result, some impacts can be prepared for even when the next major earthquake is unknown. For example, high exposure and subsequent disruption to State Highway 2 (Ngauranga to Petone) and the coastal route from Seaview to Eastbourne was observed across all 10 earthquake scenarios, with severe disruption to access routes for the emergency water station in Days Bay, in Lower Hutt, suggested in five of the 10 scenarios. Conversely, the location and extent of impacts to State Highway 1 and 58 varied across the selected ensemble, and severe disruption to emergency water stations in Huntleigh Park and Karori was limited to only scenarios on the Ohariu and Wellington Faults. The distinction between consistent and scenario-specific impacts is made to indicate the variability of these impacts across earthquake scenarios, and to highlight the level of uncertainty in the manifestation of impacts depending on the magnitude and location of fault rupture. What the scenario model ensemble has stressed is the extent to which some impacts occur, irrespective of which plausible, major earthquake is utilised in impact modelling. Understanding the specificity of coseismic impacts across the GWR provides a much-needed platform to assess the criticality of risk reduction across infrastructure systems, to ensure planning efforts will mitigate impacts for a range of plausible seismic events.
  • ItemOpen Access
    Investigating the effect of outliers and laterality on knowledge-based planning optimisation models for volumetric modulated arc radiotherapy of the lung.
    (2024) Quick, Kevin
    Using a commercial knowledge-based planning (KBP) DVH estimation tool, this study aims to investigate dosimetric outcomes between multiple in-house trained KBP models for lung VMAT radiotherapy plans considering laterality and outliers within each model. This study was conducted using planning data from Waikato Regional Cancer Centre (New Zealand) under approval from the Health New Zealand Te Whatu Ora Waikato research office and Waikato Māori research committee. Six in-house KBP models were developed from 100 clinical VMAT plans with different PTV locations and prescriptions. An initial model was trained using all plans, becoming the ‘outliers- included’ model. A copy of this model was made in which vendor analytical tools were compared to an in-house developed Python script in order to judge and remove potential outliers, resulting in another ‘outliers-removed’ model. This process was repeated specifically for left-lung PTVs and the right-lung PTVs within the original training set, resulting in a total of six individual models. These models were applied retrospectively to a different cohort of 15 VMAT plans where dosimetric outcomes were compared between each model and to the original clinical plan using the same field geometry. All models performed similarly in OAR dose metrics. Interestingly, the left-lung trained models with and without outliers produced better PTV coverage regardless of the PTV laterality. Statistically significant differences (p < 0.05) between clinical and model-generated plans were observed for all PTV D₂% metrics.PTV homogeneity index was better in all outlier-removed models. Spinal cord Dmax was exceeded on one plan for both left-lung models (in a left-sided plan) despite having a lower mean than combined lung models. The removal of outliers within the training set of a KBP DVH estimation model showed little effect on final OAR dose metrics, however helped increase PTV coverage and minimise cold/hotspots. Laterality-specific models did not yield conclusive differences, although the left lung models generally produced superior target coverage compared to the right lung models in all cases. A combined lung model with outliers removed and refined optimisation parameters, mainly for spinal cord Dmax, oesophagus Dmax and PTV D₂% is the most likely candidate for future clinical use in planning automation at Waikato Regional Cancer Centre.
  • ItemOpen Access
    Predictive validity of violence risk scale – sexual offense version risk and change scores for Māori and non-Māori individuals with a history of sexual offending.
    (2024) Foster, Aylish K.
    Whilst empirical literature has supported the predictive validity of Violence Risk Scale – Sexual Offence Version (VRS-SO) risk and change scores across normative samples, there exists limited cross-cultural research establishing the predictive validity of the VRS-SO with Indigenous populations. In Aotearoa New Zealand, this limitation raises concerns regarding the applicability of the VRS-SO with people of Māori descent who have a history of sexual offending. This study investigated the predictive validity of the VRS-SO to predict sexual recidivism for a sample of 282 adult males who completed a prison-based treatment programme for people who have sexually offended against children. In total, the sample comprised 70 Māori and 212 non-Māori individuals. Overall, Māori men were rated as higher risk and had higher rates of sexual reoffending within the first 5 years post-offence in comparison to non-Māori men, however there were no significant differences in treatment- related change between groups. While mixed predictive validity was identified for VRS-SO measures across Māori and non-Māori subgroups over fixed 5- and 10-year follow-ups, most measures demonstrated similar levels of predictive accuracy across the two ethnic groups. Furthermore, Cox regression survival analyses demonstrated that Māori group membership was not associated with higher observed rates of sexual recidivism after accounting for risk and change score information. The results generally support the predictive properties of VRS- SO risk and change scores with treated Māori individuals with a history of sexual offending.
  • ItemOpen Access
    Treatment needs and change among minor-attracted and non minor-attracted child sexual offenders
    (2024) Britt, Francesca
    Child sexual abuse poses a significant social and public health issue worldwide, with the immediate and long-term effects for victims often being both severe and long-reaching. Effective treatment programmes for perpetrators of sexual abuse are crucial for reducing the likelihood of sexual reoffending and further victimisation. The term ‘Minor-attracted Person (sometimes referred to as ‘MAP’) is a relatively new term used to characterise individuals who are sexually oriented towards one or more of the following chronophilia’s (variations in sexual age interest): nepiophilia (attraction to toddlers), paedophilia (attraction to pre- pubescent children), hebephilia (attraction to pubescent children) or ephebophilia (attraction post-pubescent youth who are not yet sexually mature;) (Lievesley & Harper, 2021). Previous research indicates that minor-attracted offenders are likely to have pre-treatment needs in most treatment variables and be more difficult to treat (Beech & Ford, 2006). However, there is limited research to date on whether treatment needs to be modified for minor-attracted offenders. This study contributes to the literature on the treatment of minor-attracted persons who have perpetrated sexual abuse, focusing on how they differ from non minor-attracted sexual offenders in terms of criminogenic needs and treatment outcomes. Data from 522 men who attended prison-based child sexual offender treatment programmes in Aotearoa New Zealand were used in the current study. Minor attraction among these men was identified by their scores on the dynamic risk item D17 Deviant Sexual Preference of the Violence Risk Scale Sexual Offence Version (VRS:SO) to separate the men into minor-attracted (n = 226) and non minor-attracted (n = 293) subgroups. These subgroups were then compared in terms of pre-treatment criminogenic needs as measured by the Violence Risk Scale Sexual Offence Version (VRS:SO), and patterns in treatment change as measured by comparing pre- and post-treatment change measured using the VRS:SO. The findings indicated that minor-attracted offenders had significantly higher pre-treatment needs and risk levels compared to non minor-attracted offenders. Additionally, minor-attracted offenders exhibited significantly lower levels of treatment progress in certain areas, including offence planning and sexual compulsivity. However, there were no significant overall differences in treatment progress between the two groups once pre-treatment levels were controlled for. These findings suggest specialist child sexual offending treatment programmes are meeting the needs for both minor-attracted and non minor-attracted sexual offenders but underscore the need for tailored interventions that address the specific risk and need factors of minor-attracted offenders.
  • ItemOpen Access
    Identifying the Beckenham Community’s Aspirations for the Ōpāwaho-Heathcote Riparian Land
    (University of Canterbury, 2024) Helms, Harriett; Newport, Erykah; Rhind, Sophie
    Objective: Identify the Beckenham community’s connection and aspirations for the Ōpāwaho Heathcote riparian land. Methodology: • Mixed-methods approach. • Qualitative data collected through semi-structured interviews with community group leaders. • Quantitative data gathered from a survey distributed to the wider Beckenham community. • Conducted various water quality tests to assess the river's health. Key Findings: • Strong community support for native riparian bush regeneration. • Interest in implementing ecological education in local schools. • Desire for infrastructure developments. • These themes reflect the community's aspirations for enhanced care and comfort at the Ōpāwaho-Heathcote River. • The area is frequently visited by respondents, underscoring its importance for community well-being and connection. Limitations: • Limited timeframe restricted the analysis of additional research that could have benefited the project. Future Research Recommendations: • Create a visual plan of the riparian land that incorporates community aspirations. • Conduct further water quality tests that account for external factors. • Analyze soil composition and structure to support successful vegetation growth.