Introduction

In countermeasures to increasing globalisation and efforts to reduce human impact on the global systems such as the lithosphere, biosphere, hydrosphere and atmosphere, alternative food networks (AFNs) provide localised solutions to limiting the contribution of conventional agri-food systems to global warming and environmental degradation [1]. Urban agriculture (UA) and other community-centric initiatives are increasing in popularity due to their associated benefits including reduction of living costs, ecosystem benefits, self-sufficiency and general wellbeing [2]. This review aimed to identify the different aspects of AFNs which achieve these benefits and promote circular economy within the urban food system (UFS).

Context

The circular economy is a system which aims to separate the human economy from its reliance on non-renewable resources, placing emphasis on restorative instead of wasteful industry models and ceasing dependence of economic growth on use of virgin materials [3]. As a tool for achieving sustainability goals [4], the aim of circular economy is the transformation of the current linear economy to a more circular conceptualisation. An extensive body of literature regarding circular economy exists, with more focus on engineering and natural science solutions than economic and management solutions to issues with scopes ranging from micro (firm level; products) to meso (cooperation between firms; resource loops) to macro (city, region or national levels; life cycles) levels [4, 5].

Of significance to circular economy is the agri-food industry because of its presence interspersing many human systems and its current unsustainable structure. The current linear model describes the agri-food system as the production, processing, distribution and consumption of food, throughout which different waste streams are outputs, revealing the inherent unsustainability of the system [6]. Of relevance to this review, Oceania’s agri-food system is unsustainable because of the energy imbalance in its production, where the caloric value of food produced is 0.76 kJ per kilojoule of fossil fuel input [7]. This is in part because of the amount of food waste experienced in Australia. The Australian Government, in their National Food Waste Strategy, has set a goal to halve national food waste by 2030, with priorities to target improvements in policy, business, market development and behaviour change [8]. These issues illustrate the significance of the issue within the Australian agri-food system and depict an agri-food system that could benefit from transition to a circular economy. If the concept of a circular economy is to become a future reality, the identification of materials, practices and structures enabling this transition is essential to their functional implementation within the agri-food sector.

In Australia, development of a circular economy within the agri-food industry has begun. The focus of development has been within the UFS, where AFNs which promote sustainability on micro and meso levels, the nature of which are often contrary to the conventional unsustainable agri-food system. AFNs are defined as agri-food system participators which function within the conventional system and are characterised as sustainable participators [9]. Because of Australia’s unique geography, the current predominant UFS is characterised by distinctly long supply chains, rural production and centralised food distribution [10, 11]. One of the main advantages of AFNs is the achievement of greater resilience and sustainability due to the shortening of food supply chains [12, 13], which is facilitated through various local food production methods including UA. In Fig. 1, the study scope is shown to be the meeting between the boundaries of the UFS and AFNs where circular economy is present. The necessity of specifying the study scope to this level is to filter and identify materials, practices and structures which are involved with the creation of a circular economy within AFNs situated in the UFS. As can be seen, AFNs also exist outside of the urban space, but obviously within the agri-food industry. Due to the geographical containment of UA to within urban spaces, UA is completely within the UFS. The scope of study encapsulates aspects of all three of the UFS, AFNs and UA, but also excludes aspects of them which do not promote circularity. Clearly, the size of each system sphere in Fig. 1 is not representative of their scale.

Fig. 1
figure 1

Venn diagram depicting the classifications of phenomena at levels within the agri-food industry and the study scope

Full size image

This review argues that AFNs promote aspects of circular economy in their achievement of sustainability. Because of the wide scope of the agri-food industry, AFNs represent a fraction of the literature regarding sustainability in their industry and further analysis of circularity within AFNs is also sparse due to its definition firstly as a sustainable practice. There is a gap in the current literature in the linking of engineering and science research on individual technologies, methods and processes to higher-level management processes, as identified by the need for an interdisciplinary approach to enable implementation of circular economy within this industry [4, 6]. This review therefore aimed to systematically identify the factors of AFNs contributing to circular economy and how they interact with the food-energy-water (FEW) nexus. While this review did not intend to formulate management strategies, it was intended to encourage interdisciplinary participation by identifying functions of AFNs contributing to circularity within engineering and scientific literature to better inform higher-level management plans. In doing so, synergies between resources, practices and management could be identified to facilitate the creation of management strategies and to highlight micro-level phenomena that enhance circularity and sustainability to implement in the broader UFS.

Study scope

This systematic review examined the capabilities of current practices and technologies of AFNs within the Australian UFS to promote circularity and sustainability. This review aimed to collect current knowledge on AFN management strategies, practices, technologies and resource compositions of flows within the FEW nexus, with a view to identify synergies between these and to highlight their benefits so that management strategies could be best informed. The following research questions were used to direct a systematic literature search:

  1. 1.

    What are current practices and technologies within AFNs that provide positive externalities of sustainability, ecosystem services and cleaner production towards circularity?

  2. 2.

    How do these practices and technologies interact with and enable the transformation of the food-energy-water nexus into a circular model?

These research questions were formed to conceptualise the multifaceted complexity of AFNs into three overarching contributions: material, practical and structural. These themes, defined in Table 1, provided a unique means of perceiving the FEW nexus as a collection of tangible and intangible elements. The study’s shift from conventional perception of the FEW nexus beyond the material adds value to the existent body of literature by providing perspective. Hence, the review was structured in a systematic method and analysis was guided by the pillars of AFNs defined in Table 1. Finally, the review was concluded by drawing connections between engineering and science solutions within AFNs which could be utilised in management strategies to best capture the value of synergies between phenomena influencing the FEW nexus.

Table 1 Definitions of factors contributing to an alternative conceptualisation of the AFNs
Full size table

To conceptualise how these factors contribute to and influence the FEW nexus within AFNs, the interactions between these three themes were illustrated in a flow diagram. Fig. 2 depicts AFNs as the connections between the producer and the consumer, both physical and functional. Physical flows of materials and energy circulate between participators, with transformation of products occurring to provide value to each participator. Transformation of materials and energy to the consumer manifests itself in forms such as on-farm production and supply chain distribution, while the same flow to producers is subject to transformation such as via the lithosphere, biosphere and hydrosphere to return to producers the raw resources. Waste flows within each step of the value chain occur due to inefficiencies inherent to the system such as nutrient loss and food waste. Practices and behaviours of participators within AFNs influence other participators’ interactions and interactions of the network on the material and energy flows. Similarly, structural elements of AFNs have a higher-level influence on how flows within the FEW nexus are treated. Their wide scope allows large influence on systemic function within AFNs.

Fig. 2
figure 2

Representation of the physical and functional flows interaction within AFNs

Full size image

Methods

This review followed a systematic approach, yielding a set of results from a keyword search in two databases and a grey literature search. Scopus and Web of Science were the two databases used to obtain relevant documents, primarily journal articles, with review articles and a conference paper also utilised. Due to Australia’s unique geographical location, the review was limited by requirement of the research questions, which limited the study scope to domestic urban Australian agri-food and supply chain systems.

Although this study focuses on flows contributing to the (FEW) nexus of the UFS, the term ‘food-energy-water nexus’ was not included as a search term. Instead, the terms ‘practice’, ‘technology’ and ‘resource’ were used because they were more indicative of the research questions and returned results more relevant to actions resulting in flows as part of this nexus. By taking a process-based approach rather than a resource-based approach, the review aimed to identify practices, technologies, strategies, and resource flows and collect them into a portfolio ordered as per the conceptualisation in Table 1. The keyword search within academic databases was structured to cast a wide net over all aspects of urban AFNs. The search string was developed incrementally by refining these research intentions into keywords and is quoted here: (“urban agriculture” OR agroecology OR agrifood OR agri-food OR “urban food” OR “community garden” OR “urban farm” OR “alternative food” OR “farmers market”) AND (production OR consumption OR distribution OR “supply chain” OR network OR system OR technology OR practice OR resource) AND Australia).

Search engine refinements limited the search to publications between 2010 and 2020. Exclusion of irrelevant academic fields including social sciences and medical studies distinguished the subject area to geography, agricultural and biological sciences, environmental sciences and ecology, earth and planetary sciences, energy and water resources, urban studies, food, plant, soil and nutrition studies and engineering. A literature ranking was employed to limit the body of literature to peer reviewed journal articles, conference proceedings, reports, case studies and surveys in that order of importance, with lower-ranking studies only included if unique conclusions were reported [14]. Documents were first manually filtered by assessing relevance based on their abstract. Afterwards, once duplicates were removed, articles were finally filtered based on the relevance of their content to the research questions. Fig. 3 outlines this process used to search and filter documents, with the resultant number of documents listed at each stage.

Fig. 3
figure 3

Database search refinement process for systematic review, with document pool listed at each stage

Full size image

Due to the incremental development of the search string, a cut-off criterion was used to limit the query specificity, necessary to retain diversity of sources. The cut-off criterion for ceasing the academic literature search was discovering fewer than five additional relevant documents across both databases when exclusionary keywords were added to the string. With the addition of search terms ‘practice’, ‘technology’ and ‘resource’ to the second string, a total of 17 additional sources were found, with two assessed as relevant based on the content of their abstract. Therefore, the set of relevant academic literature was then regarded as complete, and a subsequent grey literature search was conducted. This grey literature search included filtering of government reports such as from CSIRO and searches from databases of Australian NGOs embedded within the agri-food industry. The grey literature search recruited another four documents, which concluded the literature search at a total of 46 relevant sources (Fig. 3).

Material and energy flows

Material flows

Plastic use in distribution chains

Plastic use is inevitably driven by quality assurance and convenience, despite the disentanglement of AFNs from the neo-liberal, productivist food system. Two empirical surveys conducted within a farmers’ market and a box scheme in Brisbane revealed that, despite intentions to the contrary, plastic persisted in the form of delivery boxes, liners, bottles and bags [15, 16]. The reasons for their continued use was because they were ‘the best solution’ to fulfil their functions to preserve quality, meet hygiene standards, mitigate weather effects and increase prosumer convenience [16]. Although some plastics such as liners, boxes and bottles were returned, the process of collection and sterilisation was more expensive to time and labour. Furthermore, reuse of plastics only delayed their inevitable disposal due to fatigue and damage incurred over consecutive reuses, such as the liners, which only lasted 4–5 reuses [15]. The conflicting requirement of plastics to fulfil certain tasks and the motivation to reduce plastic waste is an inhibitor of circularity in retail AFNs. A case study comparing Brisbane and Melbourne AFNs found similarities in waste reduction measures, in which specialist retailers, food hubs and buyers’ groups were able to minimise plastic waste through pre-ordering, bulk-buying and minimising packaging [17]. However, plastic could not be eliminated. This common motivation in all studies to reduce impacts has evidently led to similar measures to reduce plastic waste, but also similar limitations, which have not been able to be overcome. Evidence of alternative solutions includes the use of cabbage leaves as clamshell containers for berries, and paper liners instead of plastic liners [15]. These solutions are met with limitations including availability of cabbage leaves, wet conditions disintegrating lining and poorer performance in retaining freshness and quality in transportation. Although alternative solutions exist, their adoption is restricted by limitations in quality assurance, cost and reliability conflicting with the environmental benefit of avoiding plastics in a fundamentally economic problem. The superiority and utility of plastics to overcome these limitations necessitates that, until an equally performing substitute is designed, plastic and plastic waste will continue to permeate AFNs.

Soil as an intrinsic medium in urban agriculture

A reciprocal relationship between soil health and urban agriculture makes it an essential medium for nutrient transmission; however, lack of proper maintenance causes external impacts. An educational course run internationally and in Sydney argues that phosphorous use in fertilisers presents opportunities for yield efficiency gains, recycling and integration into urban farms, but it also presents risks to terrestrial and aquatic health [18]. Similarly, overapplication of fertilisers was found to lead to runoff and a subsequent decline in the health of waterways, as reported from a study in the Peel-Harvey catchment in Western Australia. However, only 5.6% of survey respondents considered environmental impacts of fertiliser application when purchasing fertiliser [19]. The disparity between environmental effects and consideration of impact and the irrespective poor sustainability of fertiliser and phosphorous use reveals the need to educate consumers in correct use of synthetic fertilisers. This issue extends to pesticide use, which is a detrimental waste flow also affecting aquatic biodiversity in the Hawkesbury-Nepean River catchment on the peri-urban fringe of Sydney [20]. As with fertiliser use in [19], in this case there is no current regulation of private pesticide use or monitoring of river health, both of which have been identified as major issues. These problems are highlighted in the large environmental impact of small, highly diversified farms ubiquitous to the region [20]. Due to the lack of urgency to regulate pesticide and fertiliser use in these regions, this is of serious threat to the holistic sustainability of UA, due to offsite implications. The necessity of soil amendments to promote agricultural productivity must be met with enough production of low impact alternative products and adoption of proactive sustainable soil health maintenance.

Contrasting to the studies in the Hawkesbury-Nepean and Peel-Harvey catchments, evidence suggests that soil health and sustainability are both achievable in UA. A study of 12 community gardens within 15 km of Adelaide reported conditions promoting soil health and high site biodiversity. This was achieved using raised beds to promote cleaner production and avoid metal-contaminated indigenous soils, and derivation of most plant nutrients from urban waste streams [21]. Likewise, a study of 13 vegetable gardens within 10 km of Melbourne CBD also identified raised beds as having lower lead concentrations. In this study, most samples had metal concentrations below Australian residential health investigation levels, except for lead in 8% of community gardens and 21% of residential gardens [22]. Furthermore, to ensure limited metal uptake in edible portions of vegetables, pairing biofilters with hydroxyapatite or lime soil amendments reduced lead concentrations from irrigated stormwater, transmitted through soil [23]. Since differences may exist between prevailing lead in soil and transmission from stormwater irrigation, this potential solution requires greater research before determining the method’s viability. Historical and diffuse legacy contamination of lead in soils was in lower concentrations in residential gardens, analogous to the findings of the Adelaide study whereby historical industrial sites had higher concentrations of copper, lead, nickel and zinc [21, 22]. These two studies showed that, despite impacts of historical site use and limits to applicability of solutions, sustainable soil health in community gardens is possible, but application of measures is case-specific. Soil health in UA is thus achievable due to measures such as utilisation of urban waste streams for organic compost and use of farming methods receptive to existing soil contaminants.

Utilities in urban agriculture

Water use in urban agriculture

A key enabling factor of urban agriculture is access to reliable water resources. As seen in the Millennium Drought, for Australian cities experiencing summer-dominant water stress, irrigation using alternative water sources is a means of reducing net water use [24]. With most of Australia forecast to become drier due to climate change, exacerbated by a future doubling in frequency of extreme El Niño events [25], reuse of treated and untreated greywater, stormwater and wastewater is a viable means of protecting water security. Wastewater collected by rooftop rainwater harvesting, sewage treatment and stormwater treatment are various means of reducing net water use and promoting circularity in UA [23, 26,27,28,29], especially in water-scarce cities such as Adelaide [13]. Furthermore, water security threats of climate change and climate variability ensure that wastewater use acts as a protective buffer for agricultural productivity Australia-wide [27].

Water-efficient irrigation systems service home gardens and some community gardens, however supply variances in different cities ensure that uptake is not universal. Across the country, water use efficiency (WUE) varies due to a wide range of measurement variables including roof catchment areas, tank sizes, tank lifespans, rainwater harvest yields and water use habits [30]. Consequently, there is no clear consensus on an optimal means of increasing WUE in UA. Additionally, various qualitative and quantitative metrics for WUE’s effect on increased productivity exist in literature. Different means for increasing productivity are implementable, and expression of each method’s productivity varies. For example, one study found a 20% increase in yield while irrigating a 20 m2 garden with harvested rainwater, while another found that increased irrigation enabled by water-efficient systems subsequently increased yield, which was also viewed to decrease the food’s overall carbon footprint [26, 31]. This inconsistency in reporting of results problematises comparison of WUE in different studies. Pollard et al. argues that this necessitates future consistent reporting of rainwater and irrigation along with disaggregated yields per crop type as a minimum standard for consistency [30]. The agreement within these studies that greater total water used in irrigation positively correlates with greater yields was agreed upon by Csortan et al. [32]. This South Australian study from November 2016–June 2018 was conducted through a citizen science program called Edible Gardens [33] and provided useful insights into increasing water efficiency in UA. Although there was much variability in yields (0.02–1.42 kg/m2/30 days), greater total water use corresponded with greater yields and higher economic value [32]. Therefore, wastewater irrigation of UA can facilitate larger crop yields, decreasing dependence on conventional food and water supply. Furthermore, various methods of increasing WUE are all shown to increase productivity, providing dual benefits to circularisation of the FEW nexus.

Multiple studies discussed the underutilisation of recycled water sources, showing that there is potential in Australian UA systems for greater efficiency. Although WUE in the Edible Gardens project, which used both municipal and wastewater sources, was comparable to that of conventional agriculture, UA has demonstrated to be able to provide greater productivity per unit area, even when farming organically [32, 34]. From this South Australian project, the total water use of involved UA initiatives was 690.8 kL, divided between reticulated mains supply (71%), harvested rainwater (30%) and bore or grey water (5%) [32]. This corresponds with Adelaide’s high wastewater use but is at odds with national wastewater use. Adelaide recycled one third of its wastewater, while the national average was only 10% in 2015 [13]. One consideration for large-scale urban wastewater management not considered in Australian studies was the use of nature-based solutions such as treatment wetlands. Such a non-chemical treatment method would implicate significant reductions in energy use to treat wastewater and allow for both nutrient and water recycling [35]. Such a system has been suggested for achieving circularity within the energy-water-environment nexus within Qatar [36] and Vienna, Austria [35], however seems more suited to decentralised management systems, not common in Australia, despite their purported benefits to circular urban and peri-urban greening initiatives. Nevertheless, urban greening including for UA has been achieved in Adelaide via chemical wastewater treatments [13]. Exclusive of Adelaide, the rest of Australia lacks initiatives to manage wastewater in large-scale circular systems, which represent uncaptured value in water recycling for UA systems.

Although the national proportion of wastewater use is low, this is not a barrier for water irrigation approaching self-sufficiency. A Gold Coast case study determined that the proportion of an eco-estate’s water use that was self-sufficient was 91%, with all wastewater treated onsite [28]. This was achieved through a combination of onsite sewage treatment facilities and water-sensitive urban design, which is a scalable design for other urban areas. Such scalable design may be the key to further improvement of Adelaide’s unparalleled wastewater use, despite it being one of the driest urban areas in Australia. State- and government-funded projects such as the Virginia Pipeline Scheme, which provides over 22 GL per annum to over 250 farms on Adelaide’s peri-urban fringe show that irrigation practices utilising wastewater, need not be constrained to household- or estate-scale harvesting efforts [13]. What is not discussed in any of the literature is the energy impact of wastewater treatment and delivery compared to that of mains water use. Overall, wastewater use in Australian cities is underutilised, but prominent examples of water sensitive urban design display transferrable capabilities to increase self-sufficiency and circular water use to decrease dependence on reticulated mains water.

Although recycled water use in UA may have significant potential, conditions for safe use exist. Passive onsite treatment of potentially unsafe stormwater using biofilters can enable adoption of roof-harvested rainwater use in UA, with some exceptions on. A lab study on soil metal contaminants supplied via stormwater used soil columns fitted with biofilters and dosed them with synthetic stormwater at rates comparable to the Melbourne climate [29]. Five sets of nine replicate soil columns each containing different vegetables found a 70% decrease in copper, lead, zinc, manganese and nickel concentrations and a decrease of 47% and 69% in nitrogen and phosphorous effluent concentrations respectively. However, cadmium and lead levels in edible portions of most vegetables was found to be above WHO guidelines for healthy consumption [29]. A subsequent study added soil amendments of lime or hydroxyapatite to the biofiltered soil column and produced encouraging results for confidence in stormwater use for UA. Although only three vegetables in 15 soil columns were studied, all vegetables were found to have cadmium and lead concentrations healthy for consumption, except for radishes [23]. This result was attributed to root vegetables being in contact with the filter media [23]. Biofilter technology paired with soil amendments and informed consideration of vegetable selection is thus a viable strategy for safe use of stormwater irrigation. While design to ensure safe metal concentrations is essential when irrigating crops with stormwater, untreated greywater use is not recommended due to the heightened risk of infection from enteric viruses, even with treatment [24]. Circular economies requiring participation from commercial UA businesses may thus be limited in choices for recycled water use, with health guidelines permitting stormwater, constrained by crop type and method, but not permitting untreated greywater use. Nevertheless, under the correct conditions, using emergent technologies, stormwater irrigation may have an important function in the harnessing of wastewater flows for both commercial and non-commercial facets of UA.

Use of digital technology and management innovations are methods which currently improve WUE in UA and have scope for wider adoption. Automation of irrigation to increase efficiency can be achieved through Uplink and Downlink programs and incorporation of the Internet of Things (IoT) [37]. The use of these innovations coupled with crop databases provide unique digitalisation through emergent technologies such as closed-loop irrigation, which accounts for water budget, timing schedules, soil type, farm management and climate, [37, 38]. Although this system is better suited to conventional industrial agriculture, the versatility of the program enables potential use in UA, because of its incorporation of practical limitations into scheduling and scale [37]. A Melbourne study on multifunctionality within UA identifies technology to enable efficiency of other farm processes such as wastewater use through the innovation of ICT [39]. Multifunctional use of wastewater including in UA can not only provide greater efficiencies; accompanied with risk management practices, but it can also reduce water waste and provide ecosystem services. For example, treated sewage water and stormwater treated in a ‘sodic swamp’ is used on the Western Sydney University’s Hawkesbury campus to manage surrounding risks associated with fire and landscape hydrological function, while also helping sustain agricultural productivity in peri-urban land [40]. Positive externalities resulting from the employment of this holistic management system include bushfire mitigation, soil health and associated ecosystem services, and contribution to biodiversity [40]. Therefore, through informed wastewater management, modernised irrigation systems and multifunctional water use, efficient water use for UA can be realised.

Energy and labour in urban agriculture

Use of products with lower embodied energy is a means of decreasing absolute energy use in UA and often necessitates more circular use of materials. Decreasing the carbon footprints of water transportation via wastewater harvesting, use of organic waste instead of mineral fertilisers and multifunctional agriculture are all proven methods for reducing energy demand in UA [12, 21, 39, 41]. Input-output analysis as a means of identifying such energy efficiency increases has shown its applicability in UA. An extensive input-output analysis of 13 organic farms and gardens in Sydney and Wollongong devised an improvement from a 14.66% proportion of renewable inputs to 43.03% under maximum substitution for renewable inputs [34]. This was achieved by modelling self-production of compost, seedlings and water. Using mean embodied energy loading ratio (ELR) as a metric for mean energy efficiency of the gardens, a subsequent reduction of mean ELR from 5.82 to 1.32 eventuated, within the range of a sustainable system (ELR under 2.0) [34]. It is clear from these studies that a defining initiative for decreasing energy inputs and outputs is increasing self-sufficiency through circular use of materials. An application self-sufficiency in an alternative method of urban food production, in aquaponics, supports this notion. An energy and water self-sufficient aquaponics system in Melbourne has demonstrated that, despite relatively higher energy inputs associated with this form of UA, through innovation and low-impact design of water and air flows, onsite renewable energy is utilised to a high level of efficiency [42]. Reduction in energy demand and embodied energy in UA results from increased self-sufficiency, namely incorporation of more sustainable and circular practices and renewable technologies.

Agriculture in the urban space is dominated by manual labour, and research into efficiency gains in common energy-intensive tasks yields differing results. A contrasting theme regarding land use and energy efficiency was apparent in both individual gardens and when extrapolating to the complete network of UA. The Edible Gardens project in South Australia found an inverse relation between area and most metrics used to measure performance, including labour [32]. Labour is a form of energy expenditure and is not usually perceived as an energy flow, however in the context of UA, this is a primary energy flow. When upscaled to encapsulate potential citywide incorporation of UA, the energy savings compound. The reduction in GHG emissions from increased backyard use for UA was 3.23%, 5.25% and 7.26% for land use of 25%, 50% and 75% UA [12]. While this linear regression model is speculative and optimistic, it does provide a scalable objective for UA in Sydney. The major limitation of this model is the assumed area coverage of labour, which was higher than in some rural labour studies. The survey to build the model, based on 20 urban farms, also provided evidence for increasing area requiring greater energy expenditure through labour per unit area [12]. These two studies are in direct contrast with one another over their area energy efficiency results, using labour as a quantifying metric. This may be in part due to the differing approach used in data collection, via surveys and citizen science. However, a more telling difference is that the Edible Gardens survey only witnessed significant labour reduction in garden areas greater than 100 m2, with much variation before then, while the Sydney survey’s largest area was 69 m2, and experienced greater variation in energy efficiency with larger areas [12, 32]. Consequently, closer study involving more reliable data sources is required before a conclusion can be made regarding the intensiveness of urban farms in relation to area. The identification of labour as a key energy input into UA systems is nevertheless important when considering the energy impact of farm area.

Practical and behavioural elements

Benefits of permaculture in urban agriculture

Permaculture is a low-impact farming practice with sustainability, biodiversity and ecosystem services positive externalities of its incorporation in UA. As a well-subscribed method of UA, permaculture farming methods were adopted by 48% of 50 urban farms in Brisbane and the Gold Coast [43]. Gardens in this study made use of various practices to lower the environmental impact of their gardens, including use of compost bins (92%), worm farms (80%), being chemical free or organic (86%) and creating original soil (76%) [43]. This subscription rate to low-impact and self-sufficient practices is therefore not limited strictly to permaculture gardens, with many other urban farms practicing low-impact methods. A Melbourne study found that 11 of 15 interviewees used permaculture as their contribution to transformation of the food production system through revitalisation of soil, conversion of water and food waste into a resource, and climate change mitigation [44]. These circular benefits of permaculture were achieved through low energy inputs and localisation of the UFS [44]. Permaculture farms such as ‘forest gardens’ or urban food forests in Melbourne are seemingly ingrained into the fabric of urban AFNs. Not only this, but they provide space for efficiency gains in water use and productivity [45]. However, these were not the only gardens studies which aimed to protect food security in Melbourne, because all the farms utilised compost, among other diverse low-impact farming methods [45]. Clearly, most urban farmers adopt permaculture practices because permaculture represents an environmentally benign form of UA. However, this is not always the case, as some permaculture farms use and inadvertently propagate weeds. As experienced in Brisbane and the Gold Coast, cultivation of invasive species such as blackberries for the purposes of nitrogen-fixing of soil can lead to environmental impacts through their proliferation [43]. Although this weed may be out of place, blackberries do provide food and challenge the prevailing mainstream view of weeds through their re-categorisation as ‘waste-as-resource’ [46]. Additionally, permaculture methods can be selectively practiced, enabling exclusion of weeds from permaculture gardens [43]. Consequently, permaculture is not the perfect solution for circularity within UA, but the practice is valuable to a circular food network through its low-impact productivity, despite the possibility of weed proliferation.

Technologies to promote interconnections

The incorporation of networking technologies into the UFS is required for more efficient links between aspects of the supply chain. For agriculture to retain its foothold in peri-urban areas of Melbourne, multifunctionality is touted as an enhancement to compete against urban infill. Multifunctional agriculture encompasses the addition of value to agricultural practices through production of food and other services, which contributes both socially and ecologically to the spaces where they situate [39]. Clear economic benefits of multifunctional agriculture were identified as sound ecological practices, reduced production costs and access to niche markets. In these cases, technology can be used to bolster the efficiency of practices, such as in the IoT [37, 47]. Digitalisation of the supply chain is important with such a revolution in agricultural practices. Emergence of online AFNs such as online shopping as an alternative avenue to connect with customers has resulted in changing definitions of value provision, with greater emphasis on cost minimisation and product compliance as quality attributes [48]. These overarching trends within the industry illustrate the value in moving operations online, with potential for agriculture in the peri-urban space to become competitive with urban development, due to multifaceted and technologically aided innovations away from singular competencies of food production. Therefore, it is of great consequence that technologies are being developed to connect agricultural information networks through digitalisation. Evidenced by the emergence of the IoT, digital efficiency and automation optimisations, such as closed-loop irrigation systems, identification of trends in the digitalisation of the industry is important to forecast areas of innovation for AFNs [37]. Diversification of agricultural knowledge, specialisation and privatisation of advice networks, commodification of agricultural data and niche gaps in the market are all trends in the market leading digitalisation [49]. Through adherence to advice through emerging networks, emergence and propagation of AFNs in peri-urban and urban spaces is made more accessible through digitalisation.

Consumption behaviour

Behavioural patterns of consumers are impacted by economics of food production, with diet influencing both cost and efficiency of UA. Through linear programming and Land Use Diet Optimisation (LUDO), comparison of the area required for different diets can be made. In northern Adelaide, the minimum practical land footprint of a vegetarian diet of 1800–2000 m2 was a tenfold decrease compared to minimum land requirement for a high meat diet [50]. Similar trends were established in a different study between LUDO of a standard meat diet and a lacto-ovo vegetarian diet, in that offsetting consumption through UA contributed more to increasing dietary value than it did to decreasing costs [51]. While these studies offered different metrics in corroborating the decreased land-use of a vegetarian diet, Ward clarified that while UA incorporation had potential for a better-quality diet, excluding meat from diet had a greater effect on cost than on UA [51]. Therefore, cost may be a factor that indirectly contributes to greater land efficiency in UA. In a similar analysis of land use, life cycle assessments comparing home-based and conventional urban chicken meat farming found that although industrialised farming had a larger carbon footprint (7.7 kg and 4 kg compared to 2.6 kg CO2eq/kg chicken meat), the land use was smaller [52]. The varying footprints of different inputs of water and feed are cause for consideration of sourcing of chicken meat products. All these studies show that using discretion when selecting food is an important factor in determining the environmental and land-use footprint of that food, especially when that food is produced in an urban space, due to the influence of diet on commercial and home-based UA. The importance of selecting low-footprint food over low land use area is shown to be paramount in achieving best practice for a low-impact diet.

Structure of urban alternative food networks

Re-localisation and food security

The current structure of the traditional UFS is geared towards industrial production and long supply chains which can contribute to food insecurity. Neglect by supermarket chains of local producers in the Mary Valley north of Brisbane has led to food insecurity because of its dependence on transport networks [11]. During the 2011 Brisbane floods, local producers excelled in supplying emergency provisions to out of reach areas through use of pre-established, highly flexible delivery modes, reinforced by trust and communication networks [10]. Despite these significant advantages in access-restricted areas, efforts to contribute to emergency provisions and charity organisations during this crisis was inhibited by a preference by relief services for larger companies such as supermarkets, institutionalised by the antecedent supply chain [10]. Small-scale peri-urban farms, at risk of urban encroachment and climate change themselves [11, 53], are dwindling in numbers in the Mary Valley due to the decrease in financial support from the neo-liberal federal government [11]. In a contradicting perspective from survey participants, localised food supply was still considered important and contributing to food security. The actions of AFNs during- and post-flood has emphasised the superior food security provided by AFNs through the resilience of shorter supply chains and transport adaptability, valued regularly and in crisis. Despite this, local urban and peri-urban food supply is still not preferred in the prevailing structure of the UFS.

Several improvements to local supply chains have been identified in learnings from the events in 2011 and beyond. These include the requirement to protect and revitalise dwindling numbers of urban farms in the Mary Valley and utilisation of rail networks to overcome supply issues and reduce overall GHGs from a road-dominant supply chain [11]. However, food security cannot be achieved through improvement of AFNs alone. The Australian Capital Territory is an example of this, where its low population density ensures a nutritional self-provisioning capability of greater than 100% [54]. Alteration of the balance of imports and exports to the ACT has potential to destabilise current economic stability of conventional supply chains. Regardless, Porter et al. casts doubt over the ability of urban centres to fully provide for themselves due to increasing population competing with increasing productivity [54]. Thus, re-localisation can aid transition towards a low-carbon UFS by acting as a buffer for partial reform of supply chains and filling the nutritional void left by food exports. To facilitate these changes, more must be done from planning and policy perspectives to protect and promote AFNs.

Management and planning of alternative food networks

Management of community gardens

More than just spaces of agricultural productivity, community gardens offer unique contributions to circular economy through shared resources, but proper management and planning is essential for optimal impact. Community gardens are a means of increasing local food supply and decreasing food miles, but they also suffer from disorganisation, leading to underutilisation of shared resources. Empirical analysis of three Sydney community gardens found that a lack of vision and biased interests in hardy or water-resilient plants due to lack of time or skill contributed to lack of diversity amongst plots [55]. Another survey-based study expressed the benefits in community garden participation influencing uptake or sustainment of behaviours such as home gardening and preference for local food systems [56]. Although inefficiencies associated with voluntary participation in community gardens exist, there is evidence that community gardens provide an opening into AFNs, whereby members become more involved and, as a result, their actions synergise with circular flows [56]. For greater involvement in and proliferation of community gardens to eventuate, garden management and planning must be recognised and implemented.

Urban planning for alternative food networks

Dedicated urban planning is instrumental in ensuring the survival and growth of AFNs within the urban space. Cases of successful urban planning resulting in positive relations between humans and the environment, such as the high level of circular self-sufficiency attained on the Gold Coast eco-estate, is enabled through a food-oriented plan [28]. Miller attributes this success to harnessing the synergies between food, energy, water and waste, implemented in planning and development stages [28]. To incorporate the FEW nexus into future sustainable urban development, Miller devised a framework which identifies important considerations for planning of wastewater treatment, rainwater harvesting, solar energy, food production and waste recycling, which all have applications in UA. The criteria for considering these systems were suitability of site, urban suitability, demand suitability and climate suitability [28]. These criteria create a replicable and measurable structure for environmentally benign urban planning; however, this planning framework has only been used in the private space. According to Egerer et al., community gardens are better suited to inner-city areas, due to stable temperatures possibly provided by the urban heat island effect [57]. In the related study of 11 community gardens in Melbourne, gardener irrigation behaviour was found to be influenced by temperature variability, but plant choice was not. Regardless of the efforts of gardeners, temperature variability was still suggested to impact on species survival and distribution [57]. Therefore, as a means of ensuring species richness in community gardens, consideration for inner-city areas is more beneficial for biodiversity, which is important for balanced food production. These considerations, paired with Miller’s suitability framework, could be adopted for urban planning of community gardens. This employment of circular, food-oriented urban design in the public space could possibly influence growth of AFNs into the mainstream.

Management and planning from a regional perspective are vital to retain and promote the value of AFNs, especially in peri-urban areas. One example of poor management of the peri-urban space by government is attributed to the decline of UA areas between more urbanised spaces and the Blue Mountains World Heritage Area (WHA). Landscape Function Analysis (LFA) was used in conjunction with a semi-quantitative life cycle assessment to show that agricultural services provided better soil stability, infiltration capacity and nutritional content as compared to urbanised surfaces [53]. Despite this, the important buffer between the fire-dependent areas within the WHA and the urban landscape has been eroded over time without protective policies in place, with the issue historically being overlooked in favour of urban development [58]. The LFA studied urban expansion in the area over a four-year period and explored options for supporting peri-urban agriculture. Proposed tools included an integrated biosystems approach to organic waste involving conversion of old apple trees to microbially support orchards sustainably and to increase farm income by mushroom production [53]. Despite these initiatives to preserve function of land essential for both fire protection and unique agricultural value, policy measures tend to oppose current landscape uses, regardless of the value that the innovations in UA provide directly to local markets [58]. There is a trending outcry Australia-wide in all types of AFNs, witnessed through literature, for planning and policy enabling promotion of farmers’ markets, community gardens, peri-urban farms and multifunctional agriculture [11, 17, 20, 39, 45, 55]. Where there has previously been support for bottom-up approaches in trends towards local food, Mason & Knowd emphasise the importance of a top-down approach due to the current inability of urban agricultural practices to remain competitive with urbanisation [58]. Consequently, governance must evolve to recognise the role of AFNs in promoting balanced social, economic and environmental quality [58]. Urban planning and policy implementation to create spaces for AFNs to thrive in Australian cities is not limited to saving the peri-urban fringe and is perhaps the greatest challenge inhibiting positive environmental change through AFNs.

Discussion of circular economy in alternative food networks

The identification of different researched material and energy flows, and practices and behaviours within AFNs in Australia provide functional implementations of micro-level contributions to circular economy and sustainability. Addressing where these phenomena fit and synergise within the greater circular economy is intended to provide value to the overall system. The literature included in this systematic review fitted the predominant distribution between engineering and natural sciences foci, and economics and management foci present in global literature between 2011 and 2021 [4]. Furthermore, the apparent lack of links, or existence of ‘superficial’ links, drawn between these engineering and managerial areas of enquiry in this review is also prevalent in much of the literature regarding circular economy and sustainability within this time frame, according to Nikolaou et al. [4]. Such superficial links between these different disciplinary approaches to circular economy is evident in the various recommendations for planning and policy to enable function and growth of AFNs in Australia in the engineering papers, as identified in section 5.2.2 [11, 17, 20, 39, 45, 55]. These recommendations identify the need for management in AFNs to accelerate circularisation, while not exploring the specific composition or specification of such policies or plans to achieve the desired outcomes. Therefore, the review exposed the already identified issue of a lack of interdisciplinary focus in circular economy research, in this case specifically in AFNs in Australia.

Structural elements of AFNs were found to be heavily influenced by the urban planning and management systems in place in the prevailing UFS. However, the potential identified in urban planning to enable infiltration of AFNs into urban spaces was found to require incorporation and use of micro-scale technical and engineering solutions. For example, planning for community gardens in urban areas can be informed by temperature distributions [57] and frameworks to ascertain suitability [28], both of which are science-based knowledge applications. Management of urban water has capacity to incorporate a range of engineering solutions through water sensitive urban design, increasing WUE [33], water use from different sources [23, 26, 29, 40] and large-scale wastewater treatment and reuse schemes [13]. Implementation of technologies to advance information networks and digitalisation of supply chains [49], and automation of farming irrigation systems [37] are changing the structure of AFNs and how they are managed. In some cases of multifunctional agriculture, economic value is being created through associated service providence in conjunction with agricultural produce [39], but the technical methods employed in farming remain constant, meaning an exclusive management perspective was employed. Such diverse examples of the potential applications of management and economic strategies implementing engineering and natural sciences technical solutions in AFNs evidences the merits of an interdisciplinary approach recommended by Nikolaou et al. [4].

The type of circular economy which is most promoted in the technical, methodological and structural aspects functioning within Australian AFNs is bottom-up sufficiency as described by Bauwens et al. [5]. Approaching elimination of linear flows and waste in the food-energy-water nexus through technologies such as those which increase WUE [26, 31], reduce plastic use [15, 16], increase reuse of water [13, 32, 34] and facilitate greater connectivity through digitalisation [49] are thematic of bottom-up sufficiency. Allowing expansion of UA through practical considerations for soil health [21, 22], diet [50,51,52], permaculture methods [43,44,45,46] and food security [10, 11, 54] are all techniques which also align with the philosophy of bottom-up sufficiency. Bottom-up sufficiency is the transition towards a circular economy supporting UFS which accommodates decentralised, small-scale production and reorientation of production towards local needs instead of for commercial gain [5]. Consequently, a consistent theme identified throughout this review is the applicability of technical solutions to promoting circular economy in local contexts. This is evident not only through research evidencing the robustness of local supply chains to food insecurity, but also in the prevalence of UA and community gardens as a vessel for circular economy values including minimisation of resource use and self-sufficiency, including behavioural changes [56]. The strong merits of localisation evident in values promoted by UA and community gardens throughout this review are supported by Liaros [7]. Liaros also argues that the current UFS structure limits opportunities for circularisation with its commercially driven preference for decentralisation and longer supply chains; a theme identified in this review.

Conclusions from thematic analysis

This systematic review focussed on the capabilities, shortcomings and externalities associated with AFNs situated within the UFS in Australia. Through this review, it was found that, although material and energy flows are direct contributors to the FEW nexus, other processes influence the system as well. As a system closely intertwined with human interaction, the effect of societal behaviour, technology and institutional practices was found to have a profound impact on the composition of flows within the UFS. What may be perceived as a simplistic alternative to the conventional agri-food network, AFNs situated within the UFS are multifaceted and nuanced systems, which are much more than the coagulation of material and energy flows through the FEW nexus.

The complexity of the structure of AFNs in the urban space necessitates a shift in the paradigm of thought surrounding the conceptualisation of the network away from a resource-based view of the FEW nexus towards a process-based view. Accordingly, this systematic review analysing the flows of the FEW nexus through AFNs in the UFS found that these flows could be grouped into three common themes affecting the balance of the FEW nexus: materials, practices and structures. Key findings from the review varied in scope and scale of their effects on the system under which AFNs function.

Conclusions from material and energy flow analysis

Material and energy flows through AFNs are essential to their function, thus such flows are discussed extensively in literature. Plastics, soils, water use, embodied energies of products and manual labour are all intrinsic material and energy flows to the function of various aspects of AFNs.

Unparalleled usefulness of plastics encourages their presence in distribution systems of AFNs, such as in farmers’ markets and as product containers. There is issue in their eventual imminent contribution to waste streams, yet alternatives are currently outperformed in some aspects of their fitness for purpose. Soil health is an important, but often neglected foundational factor in promoting productivity in UA, but sustainable soil health through low-impact methods is attainable. Utilisation of alternative soil amendments, like compost from urban waste streams, and modification of farming methods to incorporate soil health maintenance are effective low-impact methods. Urban wastewater is an underutilised resource in Australian cities and represents an opportunity for closing the water loop. Despite this underutilisation, examples of effective wastewater use are in water sensitive urban design of an eco-estate on the Gold Coast and city-wide recycled water use in Adelaide. Wastewater irrigation and increasing WUE are both employed in UA and can facilitate larger crop yields and improved productivity. Under correct conditions and with informed use of emergent technologies such as biofilters and soil amendments, stormwater use can be made safe for use in UA. Furthermore, WUE increases through informed wastewater management, modernised irrigation systems and multifunctional water use are innovations employed around Australia, not only benefiting UA systems. Overall, more self-sufficient and technology-aided water use are important contributors to circularity in UA.

Use of renewable energy and reduction of embodied energy results in increased self-sufficiency in UA. Incorporation of more sustainable and circular practices like self-production of farm inputs can decrease energy demand. Inputs such as water, fertilisers, compost and seedlings can all be self-produced to minimise embodied energy of UA enterprises. As a major energy input, manual labour has also been identified as an area for optimisation, with links made to amount of required labour and farm area. Varied results in studies of this link show the need for consistent measurable metrics for labour to better understand how labour and thus energy use in UA can become more efficient.

Conclusions from behavioural and practical analysis

The impact of behaviours and practices on AFNs is less tangible than materials and energy, but farming methods, diet and interactions within knowledge networks have broader influence on the benefits of AFNs. Permaculture, digitalisation and diet are three major influences on the practices and behaviours enacted within AFNs. The benefits of these concepts on AFNs are diverse and many.

A widespread method which implements circularity in urban food production, permaculture practices create value in soil health and ecosystem services. Although common in UA, different permaculture methods are often practiced selectively and generally promote productivity with low impacts to environment through self-sufficient, circular resource use. The main detractor to the merits of permaculture is that some systems increase the threat of invasive species spread through their use in gardens.

Multifunctional agriculture models, digitalisation and automation through the IoT are technological innovations which can directly benefit the sustainment and success of UA, especially on the peri-urban fringe. The emergence of more streamlined knowledge, product and advice networks allows for greater accessibility to AFNs, for both producers and consumers, and subsequent protection of these networks. Consumers benefiting from AFNs are also influential on their products, especially products from UA. In UA systems, intensiveness and land use are affected by consumer diet preferences. Favourable outcomes to the impact of UA occur when the selection of low-footprint food is prioritised over low-land-use-area food.

Conclusions relevant to structural function of alternative food networks

Structure of AFNs is important to how they function and how large their area of influence is. Most AFNs, including UA, community gardens, farmers’ markets and supply chains operate on a local scale. Therefore, these AFNs require greater consideration for planning and policy to aid their persistence in the urban space.

Re-localisation of supply chains, supporting shorter-ranged AFNs, can reduce the length of conventional supply chains and prevent susceptibility to failure in times of crisis. Subsequent partial reform of current supply chain configurations supporting industrial agriculture and large distribution and retail companies will require a shift in perceived value of AFNs from the perspective of the government. Facilitation of these changes requires planning and policy implementation to support local AFNs. The value of community gardens to localisation is also worthy of attention in planning and policy. Community gardens are important for fostering social and environmental connections in the urban space, providing biodiversity in inner-city regions where UA is more suited. Greater community involvement in UA has shown to encourage adoption of circular practices such as home gardening and preference for local food. An urban planning method involving considerations for biodiversity and suitability shows promise for more informed implementation of community gardens and other AFNs. In general, urban planning and policy implementation to create spaces for AFNs to thrive in Australian cities is vital for the continuation of AFNs in urban spaces and is thus pivotal in realising social and environmental benefits of AFNs.

Recommendations

Further research areas in AFNs in the urban space have been identified in this review. Research could focus on consistent reporting of labour as an energy input to UA and the feasibility of passive water treatment systems such as biofilters. Optimisation through design and modelling of holistic, low-impact, replicable urban farming methods is necessary to realise greater adoption of circular methods in UA. Design of plastic alternatives in supply chain distribution functions are engineering and life cycle issues that could benefit from further research. Finally, modelling the impacts of financial investment, urban planning or policies promoting AFNs on economic, social and environmental performance indicators would address a significant gap for realisation of the value of AFNs to supply a larger portion of the UFS. Urban planning, political and financial support also have great potential for significant progress in increasing proliferation of UA systems in urbanised areas, as was commonly mentioned throughout the reviewed literature as a major barrier inhibiting survival and exposure of AFNs. Support for an interdisciplinary approach to implementing circular economy within AFNs is a primary conclusion evident in literature. Exploring consolidation of technologies and practices at the micro scale with implementation of management strategies could extract the value of combined efforts and could support and advance circular economy in the UFS.