The concept of soil health is discussed and presented as an integrative property that reflects the capacity of soils, and specifically
of soils used for cotton production, to respond to agricultural intervention, so that it continues to support both agricultural
production and the provision of ecosystem services. Soil health is conceived as being dependent on the maintenance of four major
functions; namely: (i) carbon transformations, (ii) nutrient cycles, (iii) maintenance of soil structure, and (iv) the regulation of pests
and diseases. Measurement of individual soil properties, soil processes, functions or soil biota may not suffice to indicate the
overall state of soil health. Therefore, robust, yet simple, approaches to interpreting and measuring soil health are needed. This
work reviews and compiles a suite of pragmatic soil health indicators that may be applicable to Australian intensive production
systems. This suite of indicators may be used to measure, monitor, and report soil health in a consistent manner using a
combination of standard analytical techniques, quantitative hand-held proximal sensing and qualitative visual assessments, and
potential application of emerging technologies. 
Keywords: Earth observational datasets, Onboard technology, Proximal sensing, Remote sensing, Sensor fusion, Soil processes
and function, Spatial modelling, Soil quality.

This analysis draws on data from a subset of the Cotton Comparative Analysis (CCA) in each of the respective years, along with additional information supplemented from participants’ annual financial statements. The participant dataset includes a sample of irrigated cotton growers operating predominantly in NSW valleys and represent approximately 5% of the annual Australian irrigated cotton bale production. The data is drawn exclusively from the trading entities growing cotton. Related entities are excluded and as a result the data is limited to some extent for example with respect to debt and interest not held in the name of the trading entity. The sample represents a diverse range of grower sizes, from small-scale to large operations. Bale data from our sample and national bale data (Source: Cotton Australia, Australian Cotton Estimate) is used to scale our data to estimate industry-wide irrigated cotton data. The table below shows the number of bales produced by the sample as a percentage of total Australian irrigated cotton production.

Australian cotton and grain growers are world-renowned for producing high-yielding, high quality fibre and grains. However, there is still considerable variation in both yield and quality within and between fields, farms, and seasons. Grain quality, namely the grain protein content (GPC), and cotton fibre quality, including length and micronaire (a composite measurement of fibre fineness [diameter] and maturity), are key determinants of the prices that growers receive due to the introduction of a premium and discount system for Australian growers.
Thus, there is an onus on growers to manage for both quality and quantity to attain premium prices. Site-specific crop management (SSCM) is the practical application of precision agriculture (PA) principles, and involves the allocation of resources and agronomic practices to match spatiotemporal variability in the crop growing environment. However, uncertainty regarding the amount of within-field variation necessary to justify investment in PA technologies, and a lack of understanding regarding the drivers of this variation to support improved decision making, is a considerable limitation to the adoption of PA for growers and advisors. Today, more data is being collected on farms and by the industry than ever before (e.g. yield data, variable-rate inputs), and there is also an enormous amount of public data that is free to access (e.g. remote sensing imagery) which can be used to describe or represent variability in GPC and cotton fibre quality. By understanding how and why cotton fibre and grain quality varies within-fields, growers and advisors can be equipped with the necessary information and tools to make better management decisions for more profitable and environmentally sustainable production systems.
This thesis explores the application of on-farm and publicly-available spatial data layers for the description, characterisation, and quantification of within-field variability in cotton yield and fibre quality (length and micronaire) and GPC, and to understand the drivers of this variability within fields. Chapter 1 provides an overview and background of the Australian cotton and grains industries and the current role of PA in understanding and managing for variability in cotton fibre and grain quality. Chapter 2 presents a generalised geostatistical approach using area-to-point kriging to map and downscale areal observations of crop production data,
which is illustrated using cotton yield and fibre quality (length and micronaire) data which is measured as a module (areal/block) average. Chapter 3 demonstrates how a combination of readily-available yield, agronomic, and publicly-available data layers can be used to create a model to predict GPC within-fields to fill-in gaps in the absence of a protein sensor. Chapter 4 investigates the relationship between wheat grain yield and GPC and applies interpretive machine learning approaches using existing spatial data layers to understand the drivers of spatial variability in GPC within-fields. In Chapter 5, the opportunities for SSCM for wheat grain yield and GPC are compared by quantifying the magnitude and spatial structure of within-field variability using the Opportunity Index (OI).

While the interpretation and application of the growing plethora of spatial data layers for decision-making is a challenge for growers and advisors, this research demonstrates the how a PA approach can use these data layers to better understand the nature and drivers of within-field variability in cotton fibre and grain quality to make better management decisions for more profitable and environmentally sustainable production systems that optimise both yield and quality.

Indicators for sustainable pesticide use are essential for the cotton industry to assess, manage and demonstrate commitment to environmentally responsible cotton production. Improved pesticide management not only protects overall ecological and human health on cotton farms and in cotton catchments, but it also generates social license, facilitates market access and potentially increases the value of cotton produced.

The Australian cotton industry is committed to improving on-farm sustainability; however, as the raw material travels through the ‘value adding’ stages in the globalised fashion and textile industries, it is uncertain how its ‘sustainable value’ is transferred into the final product. The central aim of this research is to analyse how the Australian cotton industry can understand where value is created, as well as opportunities to create sustainable value along its supply chain. To explore this question, a tailored tool was developed that combines value chain analysis methods with value mapping interview techniques. This involved ‘walking’ the chain from fibre to finished fashion product to disposal. A total of 21 stakeholders were interviewed across two Australian cotton value chains from growers to retailers through to actors that collect discarded garments. Participants identified what sustainable value is, how it is created, who it benefits both in and beyond the chain (including local communities, the environment and consumers) and where future opportunities to create further value may lie. This study delivers three original contributions to the knowledge surrounding how sustainability is valued within the fashion value chain. First, the development of a method and approach which proposes another way of understanding sustainable value through ‘asking’ actors specifically what they value and why, and converging these insights to better understand the entire chain. Second, through mapping the Australian cotton value chain, it identifies actors’ experiences and perceptions of sustainability which have previously been unexamined, noting where these perceptions converge and diverge. It pinpoints the complexities that face the Australian cotton industry’s transfer of sustainable value within global value chains, such as the separation between raw material producers and retailers, as well as locked-in practices (i.e. blending fibres) which inhibit traceability and circularity. The results demonstrate a need to create a shared understanding of ‘on-farm’ sustainability. The study identifies elements to best do this through substantive (LCA data) and symbolic (storytelling) sustainability messages – and proposes how these can be co-created with stakeholders. From this, the study offers a third contribution by extending understandings around sustainability and its value within the context of fashion and textile value chains, and identifies practices that can be taken up more broadly to further sustainability within the industry.

Textile waste is a growing global issue with no effective methods of disposal currently 
available. In Australia, 85% of new clothing purchased annually ends up in landfill, which 
represents a waste of water and energy to produce these textiles. Escalating this issue, is the 
synthetic composition of these textile materials which represent 60% of current clothing 
textiles globally. To minimise the environmental impacts and relieve waste management from 
existing facilities, the concept of utilising recycled cotton fibres as reinforcement in cement 
mortar has been proposed. While most engineering practices typically use synthetic fibre 
reinforcement to obtain specific mechanical improvements, the processing and development of 
these fibres is unsustainable and environmentally damaging. 
Recent studies have found effective uses for natural fibre reinforcement in cement-based 
products. Results demonstrated comparable mechanical improvements, while minimising 
environmental impacts and costs to manufacture. In this study, cotton bed sheets obtained from 
a textile-waste recycling facility were repurposed to a fibrous state to assess the physical, 
mechanical, and microstructural interaction with a cement-mortar. A total of 7 different mix 
designs were prepared with various quantities and lengths of cotton fibres, to evaluate the 
influence of fibre content and fibre length on cement composites. 
To observe the physical properties, flow table tests were performed for each mix to determine 
the workability of fresh mix. It was observed that the increase in fibre content decreased the 
workability, and superplasticiser had minimal effect on higher quantities of fibres. Hardened 
properties were measured using compressive and flexural tests at 7, 28 and 56 day curing ages. 
The compressive strength for samples with increasing cotton fibre content demonstrated a 
decrease in compressive properties. However, flexural properties indicated a significant 
improvement in flexural toughness, and confirmed that the presence of cotton fibres was 
influencing a fibre bridging mechanism. This was supported through microstructural imaging, 
which indicated a strong correlation between fibre dispersion and the cohesion with cement 
bonding. 
Cotton fibres display promising flexural improvements in cement composites, however 
additional research should focus on the chemical pre-treatment of fibres, to enhance the 
dispersion and bonding between cement particles.

Soil is a finite and irreplaceable resource that underpins agricultural systems, regulates water cycles and presents opportunities for climate change mitigation. Consequently, the sustainable management of soil is essential for guarding the security of agricultural production and ecosystems. For this to occur, however, an accurate understanding of inherent soil properties and how they vary spatially, is required. Within the lower Murrumbidgee valley of southern NSW, Australia, there is a paucity of available data and land managers require more information on both the chemical and physical properties of their soils and the presence of potential constraints to agricultural production. While the development of novel approaches and ‘blue-sky’ research is important, the targeted application of developed methods is essential in filling knowledge gaps and allowing benefits to reach the end user, in this instance, farmers and land managers. Within the lower Murrumbidgee valley farmers are a key stakeholder of the soil resource. Thus, they should be a key consideration when designing, undertaking and communicating projects to ensure their outcomes are tangible. Therefore, rather than seeking to develop novel approaches to digital soil mapping (DSM) and digital soil assessment (DSA), this thesis utilises various methods recognised in literature to develop outputs targeted to the cotton industry in southern NSW. Using a collected dataset of 153 soil cores to 1 m depth from across the region, soil profiles are morphologically described, soil properties are modelled and mapped at the within-field and regional scales, before a regionally specific classification is developed.

Existing wireless technology and the emergence of low-cost IoT sensors in agriculture offer the possibility to gather relevant data from the soil-plant-atmosphere continuum in real time for irrigation scheduling and to have this information available remotely online. Integration of these separate technologies with automated control structures for irrigation to develop smart automated irrigation systems has the potential to reduce labour costs and increase water productivity within the cotton industry. With the final objective of developing a relatively low-cost, smart and fully automated irrigation system in collaboration with automation providers such as Padman Stops, research in this preliminary project was focused on the sensing component that such a smart automated system would have. The project showed that sensing weather, soil and crop data in real-time by means of a wireless sensor network that periodically uploads data to internet by means of low-energy and low-cost data loggers (WiField logger) offers large opportunities to optimise irrigation management through automation. In this research, data from the in-field sensors and a remotely sensed crop vegetation index (Sentinel-2 NDVI images of each bay obtained from Google Earth Engine) were ingested and analysed in Google Cloud Platform. The processed data (ETc, soil tension, etc.) plus 7-day weather forecast and soil moisture predictions were then presented in a dashboard available online using Google Data Studio in an easy to interpret manner. For the soil moisture prediction, Lasso, Decision Tree, Random Forest and Support Vector Machine modelling methods were trialled. Random Forest models gave consistently good results (mean 7-day prediction error from 8.0 to 16.9 kPa). Linear regression with two of the most important predictor variables (the square of cumulative crop evapotranspiration minus rainfall and the square of growth degree days) was not as accurate, but allowed extraction of an interpretable model. The methodology developed in this project could be used as part of a closed-loop sensing and irrigation automation system. Future research will be focused on exploring new low-cost soil and plant-based sensors, refining the algorithms developed for soil moisture prediction and the integration of such system with IoT irrigation control structures to develop a fully automated irrigation platform. With such platform, the current challenges for the widespread adoption of automated irrigation within the Australian cotton industry are likely to be significantly reduced and higher rates of adoption occur. Adoption of the proposed sensing and automation irrigation systems by cotton irrigators would result in reduced labour costs associated with irrigation water management and improvements water productivity.

The RMIT research project funded by CRDC on “The economic benefits of composting textile waste: process mapping and optimal location” focused on two distinct issues: • Identification of potential methods of transforming 100% cotton textile wastes into organic compost. • Determining the optimal location within Australia for waste processing by minimizing the transportation cost, which is reflected by minimizing the integration of cargo volume-distance between collection areas and the processing facilities, and subsequently to agricultural farms. The analysis is done through examination of existing literature and the project was divided into five operational stages as mentioned below. Stage 1 analysis demonstrates that end-of-life (EOL) consumer clothing waste in Australia can be collected via donations from households and strategic locations. Collection methods encompass direct consumer engagement, city council collaboration, and partnerships with private entities and NGOs. Additionally, pre-consumer textile and clothing waste can be acquired either directly from overseas industries or through local brokers. Stage 2 analysis confirms that textile sorting primarily depends on labels, with alternative methods like burning tests necessary when labels are absent. Sorting can be manual or automatic, the latter being more efficient through the integration of machinery, artificial intelligence (AI), and technologies like Near Infrared (NIR) spectroscopy, Fourier Transform Infrared (FTIR), and Radio Frequency Identification (RFID) sorting. Contaminants are eliminated pre-composting via industrial washing/cleaning of textile waste. Post-cleaning, textiles can be shredded manually, mechanically, or chemically. The study found automatic technologies like double shaft or Aduro shredders to be highly efficient in garment shredding. Stage 3 analysis identified three main composting methods: aerobic, anaerobic, and vermicomposting. The study found vermicomposting, utilizing earthworms, to be the most effective and nutrient-rich method for composting cotton textiles. This method yields compost higher in nitrogen, phosphorus, and potassium (NPK) compared to other counterparts. Vermicompost also retains nutrients longer than traditional compost and provides essential macro and micronutrients, including vital NPK to plants more rapidly

Riparian ecosystems support biodiversity, provide ecosystem functions and deliver various ecosystem services vital for humans (Naiman et al. 2005). Riparian vegetation is particularly important in driving the composition, structure, and function of terrestrial and aquatic systems at a local and landscape scale - influencing fish, insects, birds, mammals and soil microbiota (Capon et al. 2016). Also providing many critical services to the broader landscape such as nutrient cycling, providing fresh water and carbon sequestration. In addition, riparian vegetation dynamics are often sensitive to changed hydrological regimes, with river regulation often resulting in ecological shifts (Jansson et al., 2000).