Shaping Resilient Edible Cities: Innovative Aquaponics for Sustainable Food–Water–Energy Nexus

Qiuling Yuan; Fanxin Meng; Yingxuan Liu; Jose A. Puppim de Oliveira; Lixiao Zhang; Wenting Cai; Zhifeng Yang

doi:10.1016/j.eng.2025.01.021

Engineering ›› 2025, Vol. 50 ›› Issue (7) :257 -269. DOI: 10.1016/j.eng.2025.01.021

Research

research-article

Shaping Resilient Edible Cities: Innovative Aquaponics for Sustainable Food–Water–Energy Nexus

Author information +

History +

PDF (3651KB)

Abstract

Urban food systems need remedies due to the global rise of food insecurity. Aquaponics has emerged as a circular agricultural mode to shape edible cities and features multiple food–water–energy (FWE) nexus. Here we provide a generalizable methodology and framework to capture the FWE nexus flows of aquaponics systems within city jurisdiction. To test the framework in Beijing, China, we offer an evidence-based tradeoff analysis of urban rooftop aquaponics (RA) and ground aquaponics (GA) from a FWE nexus perspective. The results show that urban aquaponics performs well in terms of water efficiency, which saves 42%–44% of water consumption than traditional greenhouses (TG) during the on-farm stage, but generates 2.3–3.0 times higher energy consumption and 1.1–2.1 times more carbon emissions. “From farm to table” aquaponics helps decrease 14%–44% of the energy, water, and carbon impacts during the off-farm stage. With diversely optimized strategies for renewable electricity, fish food, infrastructure materials, and recycling actions, urban aquaponics can hopefully reduce energy consumption and carbon emissions by 80%–85% in the on-farm stage. In addition to greenhouse agriculture, utilizing a total of 155 km² of RA and GA potential areas could increase urban vegetable self-sufficiency by 15%, and avoid 82% of the energy, water, and carbon footprints during upstream food supply chains beyond cities. Our findings could provide policy insights for urban stakeholders to create edible landscapes by integrating RA and GA, and thus direct resilient and sustainable agricultural transformation.

Graphical abstract

Keywords

Aquaponics / Food–water–energy nexus / Rooftop agriculture / Food system / Urban resilience

Cite this article

Download citation ▾

Qiuling Yuan, Fanxin Meng, Yingxuan Liu, Jose A. Puppim de Oliveira, Lixiao Zhang, Wenting Cai, Zhifeng Yang. Shaping Resilient Edible Cities: Innovative Aquaponics for Sustainable Food–Water–Energy Nexus. Engineering, 2025, 50(7): 257-269 DOI:10.1016/j.eng.2025.01.021

登录浏览全文

4963

注册一个新账户忘记密码

1. Introduction

Cities serve as hubs where people reside and work in close proximity, favoring social engagement and shaping the abundance of food choices [1]. Urban agriculture (UA) is the practice of growing food in built-up areas and hinterlands, rather than solely associating with the imagination of the rural environment [2,3]. The Food and Agriculture Organization of the United Nations (FAO) underscores that UA can be regarded as edible green and blue infrastructures that help address various complex urban challenges, such as troubling food deserts, climate regulation, food mileage, biodiversity, and ecosystem services [4]. UA also leverages the nexus among food–water–energy (FWE) resources to shape a sustainable planet [5], [6], [7], [8], [9]. Aquaponics is appealingly among the competitive branches of UA considering resource and environmental efficiency, such as urban land, water, and fertilizer. Especially in compact cities and land-constrained areas, aquaponics actively helps improve local food sufficiency, agricultural biodiversity, and flexibility within cities, thus bolstering the resilience and sustainability of urban food systems.

Aquaponics is an innovative urban agricultural system that integrates fish farming (aquaculture) with soilless plant cultivation (hydroponics) to establish a sustainable and symbiotic interaction between fish and vegetables [10], [11]. Ground space is traditionally seen as the primary medium for agriculture, while in densely populated cities where land resources are finite and costly, rooftops represent a largely underutilized resource with significant potential for food production. Such ground aquaponics (GA) and rooftop aquaponics (RA) present several advantages over traditional agricultural practices, including reduced water consumption due to the continuous water recirculation between fish tank and plant beds, and the prevention of excessive mineral fertilizer pollution by employing fish manure as a nutrient source for plant growth [12], [13]. This mode of local UA not only helps increase multiple types of food harvest but also drives transboundary environmental footprint savings via land teleconnections [14], [15], [16]. However, aquaponics also requires much more energy consumption than traditional agriculture, because of the application of sophisticated techniques, thus indicating the tradeoffs between aquaponics and the FWE nexus.

Cities are the integration of built-up areas and hinterlands; within this lens, building rooftops inside built-up areas and protected agricultural areas beyond the city center, provide implementing space for aquaponics systems. Baganz et al. [17] identified mesoscale site potentials for aquaponics, delineated the relevant boundary conditions, and formulated policy recommendations. In addition, they took sustainable, economic, legal, and settlement structure issues into account to address the transferability of the outcomes to other metropolitan regions. Notably, one major challenge is to capture the potential tradeoffs between nutritional provisions and the negative environmental impacts associated with food production [18]. Yuan et al. [19] explored the environmental sustainability of two typical urban farms and reported that, compared with traditional greenhouses (TG), aquaponics presents a lower environmental loading ratio and higher energy sustainability index, but necessitates greater amounts of emergy inputs. Hu et al. [20] applied a life cycle approach to evaluate the carbon footprints of a conventional small householder farm and large home-delivery agriculture. The results indicated that the latter performed better in terms of the total environmental effects and economic benefits. In addition, Ghamkhar et al. [10] revealed that heat, electricity, aquafeed, and infrastructure were the major contributors to the environmental impacts of a cold-weather aquaponics system.

However, the knowledge gaps appear regarding identifying the availability of RA and GA, and further comparing these two types of aquaponics systems at different spatial scales, particularly in terms of their FWE nexus. The nexus perspective provides a systematic lens through which stakeholders can approximate aquaponics’ sustainability [21]. Furthermore, existing research has focused mostly on site-scale systems but has neglected the broader implications of aquaponics in metropolitan areas, although this would provide the future orientation of urban agricultural transformation. Based on these gaps, our research provides a general methodology and framework to capture the FWE nexus flows in both RA and GA systems within city jurisdiction. It combines the potential identification of urban aquaponics, entire life cycle environmental and economic cost evaluation, optimum design, and urban scale scenario analyses. As a developed metropolitan, Beijing has high food demand and become a national key hub for consuming agricultural products. While the limited land resources of Beijing hinder traditional agriculture’s ability to meet the growing demands, further emphasizing the need for innovative solutions. With improvements in living standards, consumers in Beijing are giving more attention to the quality, safety, nutrition, and environmental protection of agricultural products. Such consumption upgrade provides a market foundation for green agricultural models such as local aquaponics. Herein, we selected Beijing, China as a case study to validate the practicality and feasibility of the proposed framework and formulate urban aquaponics strategies for local governments. The research findings could provide policy suggestions for future agricultural transitions to increase the sustainability and resilience of urban food systems.

2. Research workflow and methods

2.1. A generalizable methodology and framework

The FWE-targeted prioritization of urban aquaponics initiatives can help decision-makers improve urban sustainability and the way in which UA orients to resilient future. In this research, a generalizable methodology and framework (Fig. 1) is established to depict the FWE nexus in urban aquaponics. The framework begins with the identification of potential areas for urban aquaponics, including both RA and GA, which provides the fundamental basis for an upscaling aquaponics analysis within city boundaries. The next step is a whole life cycle analysis of urban aquaponics based on field investigations, the assessed impact categories include energy consumption, water consumption, and carbon emissions. The life cycle stage features two distinct parts: the on-farm stage and the off-farm stage. The economic cost and profit are also computed to provide an understanding of their environmental and economic performance. A subsequent sensitivity analysis is applied to examine the pivotal sensitivity factors (SFs) of urban aquaponics. On the basis of the identified sensitivity characteristics, aquaponics systems are optimally designed to decrease the overall negative impacts. Ultimately, a city-wide scenario analysis is conducted to provide insights for future agricultural orientation; here, both local aquaponics-based agricultural development and the upstream food supply beyond the city are considered. This methodology and framework can be replicated in any city to guide the sustainability assessment of urban aquaponics and future agricultural transformation.

2.2. Urban aquaponics map

Ground areas are traditionally the primary medium for agricultural activities, and suburban regions offer ample and unobstructed spaces that are ideal for implementing GA systems. In contrast, highly urbanized areas, where land is scarce, rely on rooftops to support agricultural endeavors [22], [23]. Consequently, this research explores the potential for urban aquaponics within two specific contexts: the rooftops of buildings in densely built-up urban areas and ground areas in the suburban hinterlands. To assess potential aquaponics sites, we overlaid building typologies, sizes, and outlines. We utilized Beijing land use data with a resolution of 30 m × 30 m that encompassed 12 detailed building types—commercial, municipal, storage, industrial, public, historical, educational, residential, sport, parking, agricultural, and mixed-use [24]. These data were analyzed using ArcGIS, with type attributes assigned to each building. Various structures, including industrial, commercial, educational, mixed-use, supply and disposal, traffic area, and sport facility structures, hold promise for RA [17]. Based on our field investigation and farmer interviews, buildings with rooftop areas exceeding 1500 m² were identified as particularly suitable for aquaponics systems. Given the minimal rooftop inclinations in northern regions [25], a slope filter was not applied in this study. For GA, the research considers protected agricultural areas as indicators of potential aquaponics development. In China, protected agriculture (PA) typically includes greenhouses, large agricultural tunnels, and small and medium-sized agricultural tunnels [26]. In 2023, Beijing yielded 1.09 × 10⁶ t of vegetables within a total land area of 285.642 km² dedicated to PA. Detailed data on the development area and production are provided in Appendix A Table S1. By applying these criteria, we can effectively identify and analyze potential prospects for both urban RA and GA.

2.3. Life cycle assessment

2.3.1. Aquaponics description

The studied aquaponics system includes two interconnected agricultural tunnels that function as cohesive entities to cultivate fish and grow vegetables. This system is essentially linked by water, that is, the water with waste products from the fishing tank (aquaculture) is transported to the vegetable growing system (hydroponic). Through this flow, bacteria can facilitate the conversion of ammonia nitrogen in the water into nitrite, which is then broken down into nitrate by nitrifying bacteria (Fig. 2(a)), and the nitrate finally reaches vegetables as a type of fertilizer (see the Appendix A for more details about the principles of aquaponics system). The whole production process is devoid of pesticides and fertilizers. On the basis of the size of the urban aquaponics system surveyed through field investigations in Beijing, China (Fig. S1 in Appendix A), RA and GA were designed (Figs. 2(b) and (c)). Both the RA and GA systems consist of two agricultural tunnels. Tomatoes, lettuce, and sturgeon fish are farmed together in the first agricultural tunnel, while the second tunnel is exclusively devoted to the production of vegetables. Each tunnel, together with the area between them, has a specifically dedicated place for activities. The total area occupied by the shelters and the space between them is 1500 m². The data were obtained via in-person questionnaire surveys, data measurement, and interviews, and the gathered data encompass details on the use of resources and materials, power consumption, transportation approach, and associated expenses, outputs, and revenues.

2.3.2. Life cycle boundary

This research compared the life cycle resources, environmental impacts and the economic costs of urban RA, GA (Figs. 2 (d) and (e)), and TG (Fig. 2 (f)), which are all regarded as protected agricultural forms in this research (the details of TG are available in the Appendix A Section S1). In terms of the growing period, GA and TG allow for year-round agricultural activities, but RA is limited to growing from March to November because of the temperature limitation at relatively high elevations. Otherwise, the reduction in winter light exposure time affects plant photosynthesis, coupled with high energy costs for heating and lighting, which results in lower economic benefits for operating aquaponics systems in cold seasons. The functional unit is the impact and cost of food products per joule from the mentioned PA, during the whole life cycle stages each year. The specified inputs and outputs of RA, GA, and TG are shown in Fig. S2 in Appendix A. Here, the life cycle approach features an on-farm stage and an off-farm stage (see Tables S2–S4 in Appendix A for the detailed system boundaries). The on-farm stage includes material and construction, operation, harvest, decommissioning, and disposal for both aquaponics and TG, which differs from the off-farm stage. In the context of RA, we propose that clients have the option to buy vegetables and fish straight from the rooftops of commercial and industrial buildings, resulting in the lowest environmental footprints, and the off-farm impacts of RA is assumed as zero. For other building types in which consumers have limitations in reaching the rooftops, the products are distributed by the operator according to a predetermined sales schedule at a certain sales location, usually the ground level of the building. Residents in the area may walk to the sales site because of the proximity of RA, which further minimizes the need for packing and shipping, and the off-farm impacts remain negligible. In contrast, GA is situated in the hinterlands and the harvested food needs to be packed and transported to customers in urban areas; the transportation distance is assumed to be 39 km. Notably, however, the food from GA is transported by couriers and directly reaches consumers’ homes. In terms of TG, the food distribution process involves delivering the produced vegetables and fruits to a distribution center. From there, they are transported to downtown supermarkets via cold chains, and consumers then visit the supermarkets to purchase the goods.

2.3.3. Life cycle environmental and economic assessment

In this study, the life cycle inventories of the studied urban RA, GA, and TG were constructed based on research, literature, and reports. The background data of the inventoried substances were obtained from the Ecoinvent database, and the assessment methods of cumulative energy demand and the ReCiPe midpoint (H) were applied in openLCA software to account for the resource and environmental impacts during the life cycle of agricultural systems, including energy consumption, water consumption, and carbon emissions. The life cycle impacts can be mathematically represented as follows:

(1)

E_{q} = \sum_{l}^{L} \sum_{s}^{S} {AD}_{s} {BD}_{s, q}

where

E_{q}

denotes the life cycle resource and environmental impacts q of the agriculture system; impact category q includes energy consumption (MJ), water consumption (L), and carbon emissions (kg); l represents the life cycle stages, including the material and construction, operation, harvest, decommission, and disposal stages; L is the total number of life cycle stages;

{AD}_{s}

denotes the activity data of the life cycle inventory substance s of the agriculture system; S is the total number of substances;

{BD}_{s, q}

is the background data of life cycle inventory substance s that corresponds to impact category q of the agriculture system. Note that we subsequently quantified the energy consumption, water consumption, and carbon emissions per unit of food production (kg) per year in TG, RA, and GA by converting food production into nutritional joules.

According to the agriculture system, the life cycle cost can be divided into the initialization construction cost, future operation and maintenance cost, and waste disposal cost. As the waste disposal process involves many uncertainties, this research does not consider it but rather divides the full life cycle cost of the agriculture system into the project cost and working capital, that is, the corresponding initialized construction cost and future operation and maintenance cost. The life cycle economic cost of the agriculture system is calculated using the following equation:

(2)

EC = \sum_{z}^{Z} {EC}_{z}

where EC is the life cycle economic cost of the agriculture system (USD), and

{EC}_{z}

is the life cycle economic cost of each material and input z. Z is the total number of materials and inputs.

2.4. Sensitivity analysis

To quantitatively identify the sensitive input factors in the life cycle assessment results of urban aquaponics systems, a sensitivity analysis was performed to reveal the potential directions for aquaponics improvement. The impact of aquaponics systems varies depending on equipment selection, lifespan, maintenance plans, and other parameters. For the major input items listed in the life cycle inventory, their impacts on the current research system were calculated under ±20% variation. The relative changes in output items were compared with the relative changes in input items to calculate SFs. Input factors with SFs lower than 2% are considered negligible for the system’s impact, whereas higher SF indicate that minor changes in the input parameters could lead to significant changes in the impact results [27].

(3)

{SF}_{i} = \frac{\frac{Δ R}{R_{0}}}{{(\frac{Δ P}{P_{0}})}_{i}}

where SF_i represents the sensitivity factor of parameter i, P₀ is the initial value of parameter i, ΔP is the change in parameter i, R₀ is the result of the life cycle assessment, and ΔR is the result caused by the change in parameter i.

2.5. City-wide scenario simulation

Four scenarios are established here to explore the overall city-wide performance of aquaponics implementation, which are business-as-usual (BAU), Scenario 1, Scenario 2, and Scenario 3. As for the BAU scenario, PA includes greenhouses, large-sized agricultural tunnels, and middle- and small-sized agricultural tunnels. As for the Scenarios 1, 2, and 3, we suppose that all of the potential rooftops are used to develop the RA, as rooftop utilization does not generate additional land occupation. For GA, the entire area of PA is used for developing ① the GA, ② the TG, and ③ the integrated GA and TG. Each scenario is the same in that the potential rooftop and land areas are both utilized, but they are different in their specific development schemes (Table 1). In addition to the development areas, we further explored local vegetable self-sufficiency, which is contributed by non-PA (NPA) and PA. The remaining self-sufficiency rate needs to be met by exporting vegetables beyond city boundaries. As the scenario setting depends on various development areas of PA, there are different self-sufficiency contributions from PA. The self-sufficiency rate from local NPA is assumed to remain unchanged. Within each scenario, the life cycle performance of energy, water, and carbon can be further accounted for, and the life cycle impacts of traditional agricultural tunnels from BAU can be calculated based on the yield ratio of agricultural tunnels and the proposed TG. The economic input–output life cycle assessment (EIO-LCA) model is further used to measure the avoided transboundary environmental footprints induced by the vegetable production [11] of urban aquaponics systems, including the avoided energy footprints, the avoided water footprints, and the avoided carbon footprints.

Taking the avoided carbon footprints induced by the vegetable production benefits of aquaponics as an example, the carbon emission matrix B driven by the final demand for the avoided vegetable imports (equal to the vegetable yield) can be established based on the EIO-LCA model, which is calculated as follows:

(4)

B = R {(I - A)}^{- 1} Y

where B denotes the carbon emission matrix driven by the final demand for the avoided vegetable imports; R is the carbon emission intensity matrix; A is the direct consumption coefficient matrix; and Y is the diagonal matrix, I is the unit matrix (see the Appendix A Section S2).

3. Results

3.1. Mapping urban RA

Sixth-ring roads are regarded as proxies for urban built-up areas in Beijing, and within these areas, there are 560 000 building amounts. A total of 10.5 km² of buildings are identified as potential areas for RA, which represents a mere 0.06% of Beijing’s administrative area (Fig. S3 in Appendix A). With respect to the building categories, commercial buildings present the greatest appropriateness, covering an area of 3.4 km² with 1168 building amounts (Fig. 3 (a)). Educational buildings cover 2.3 km² with 871 building amounts, and industrial buildings cover 1.9 km² with 577 building amounts. Storage buildings show the least potential, with an area of 0.4 km² (141 building amounts), followed by sport buildings with an area of 0.3 km² (117 building amounts). Considering the municipal districts within the six-ring area of Beijing, Chaoyang District has the greatest rooftop availability with 3.21 km² of available space (Fig. 3 (b)). Haidian District follows with 2.31 km², followed by Fengtai District with 1.83 km², while the rooftop area in the remaining municipal districts is less than 1 km².

3.2. Life cycle profiles

From an on-farm stage perspective, both RA and GA consume far less water than TG when producing per joule of food (8.48 × 10⁻⁶ L for RA, 8.89 × 10⁻⁶ L for GA, and 2.04 × 10⁻⁵ L for TG), while the energy consumption and carbon emissions of aquaponics systems surpass those of TG (Fig. 4 (a)). In terms of energy consumption, RA (1.07 × 10⁻⁵ MJ) and GA (1.32 × 10⁻⁵ MJ) need 2.4 to 3.1 times more energy than TG (4.41 × 10⁻⁶ MJ). In terms of carbon emissions, RA (7.31 × 10⁻⁷ kg) is comparable with TG (6.55 × 10⁻⁷ kg), whereas GA generates 1.9–2.1 times more carbon emissions than RA and TG. This demonstrates that electricity efficiency and renewable utilization should be prioritized to support sustainable development throughout the entire life cycle of aquaponics. In addition to the different energy, water, and carbon impacts of RA and GA, their economic performance also differs. Specifically, the life cycle economic cost of TG per meter square (23 CNY) is significantly lower than that of aquaponics systems (Fig. 4 (b)). Indeed, the expense associated with electricity is one of the highest expenses in aquaponics systems [28]. Compared with RA (158 CNY) and GA (239 CNY), TG saves 85%–90% of the economic costs. However, because of the aquaponic system’s reputation for green and pollution-free food production, such high selling prices have resulted in the incomes of RA (274 CNY) and GA (412 CNY) reaching 8–12 times that of TG (34 CNY).

For both RA and GA, the operation stage is the most critical period in terms of energy consumption (70.6%–87.6%), water consumption (85.2%–89.3%), and carbon emissions (48.0%–91.0%). The main factor that influences all categories of effects in aquaponics is electricity usage (21.8%–41.3%, Fig. 4 (c)), mostly due to the need for pumping and aeration. Additionally, fish food plays a crucial role in determining the environmental consequences of aquaponics, fish food accounts for 37.4%–46.4% of energy consumption and 23.8%–45.1% of carbon emissions in the GA and RA systems. In the case of TG, 73.1% of the environmental effects can be attributed to the stage of material production and construction, and the primary contributors to this impact are brick and concrete, which account for 45.9% of the overall energy consumption during the whole lifespan of TG. The brick and concrete of the GA also account for 18.1% of the system’s total energy consumption during its lifespan, while the material stage of RA accounts for a mere 12.0% of the overall effects, with polyvinylchloride (PVC) being the primary contributor at 6.8%. Furthermore, solid waste-landfill accounts for 27.1% of the carbon emissions from GA, and those from TG increase to 45.9%. Specifically, the primary influence of water use on these three systems (TG, GA, and RA) arises from the utilization of groundwater for irrigation purposes throughout their operation. In the RA and GA systems, irrigation water accounts for 56.9% and 54.2%, respectively, of the total water consumption. However, this percentage in TG significantly reaches 93%. This also highlights the significant disparity in water consumption between both TG and aquaponics systems, and improvements in the irrigation water efficiency of PA should be considered.

In the off-farm stage, aquaponics systems clearly have significant benefits (Fig. 4 (a)). GA results in an energy use of 1.82 × 10⁻⁶ MJ, a water consumption of 1.56 × 10⁻⁶ L, and a carbon emission of 1.70 × 10⁻⁷ kg. TG is 2.3–7.4 times greater than GA, with specific values of 1.34 × 10⁻⁵ MJ, 3.55 × 10⁻⁶ L, and 8.06 × 10⁻⁷ kg. RA allows consumers to conveniently obtain their food directly from the rooftops, resulting in little off-farm effect, which can be regarded as zero. In terms of GA, vegetables and fish are delivered directly from producers to consumers’ homes via simple packaging. Compared with cold-chain transportation of TG, this technology simplifies food delivery and leads to enormous off-farm impacts; thus, significant off-farm benefits of GA are observed. However, it is evident that in both GA and TG, a majority of the energy, water, and carbon impacts during the off-farm stage can be attributed to courier and cold-chain transport procedures, which are used to preserve the products at low temperatures. This accounts for approximately 56%–94% of the overall impacts. In the case of TG, the stage of sales in supermarkets accounts for 45% of the total water use during the off-farm stage (see Appendix A Tables S5–S12 for the detailed values). Water evaporation inside vegetables must be controlled by administering water throughout the vegetable selling process. This is done to maintain the moisture content, improve visual attractiveness, and maintain the fresh taste of the vegetables.

3.3. Sensitivity analysis and scheme optimization

A sensitivity analysis provides more evidence for the key input of efficient aquaponics systems. Minor variations in electricity have a significant effect on each impact category (23%–43%, Figs. 5(a) and (b)). By including a higher proportion of renewable energy sources, the energy mix of aquaponics systems may effectively mitigate environmental consequences. However, the environmental effects also differ depending on the composition of the power grid. The environmental effects of aquaponics systems are significantly affected by the overall adjustment of energy, considering the considerable variation in electricity sources across different locations. An alternative approach to mitigate the environmental consequences of feed production is to capture fish at smaller sizes, which would result in a decrease in the overall food demand. The ultimate yield is contingent upon customer expectations, and more investigation is needed to ascertain the feasibility of the projected modifications. Research on the relationship between meat consumption and the environment has shown that it is challenging to alter people’s consumption habits based only on favorable environmental effects. Notably, the sensitivity factor for the carbon emissions of GA shifts from fish food to infrastructure materials, primarily brick and concrete.

By identifying the primary sensitivity elements, such as electricity, fish food, and infrastructure materials, we proceeded to optimize the design of the two aquaponics systems. More precisely, we examined the substitution of traditional electricity with two specific forms of sustainable energy (wind and solar power). Furthermore, we chose two types of fish food that include more ecologically sustainable ingredients (Appendix A Table S13 for the fish food compositions). The building materials used for the aquaponics on the ground were replaced with shed membranes. Furthermore, recycling rates between 30% and 70% are deemed necessary to further reduce the material burden on the infrastructure of aquaponics systems. By integrating and consolidating the optimization methodologies of the aquaponics system, we derived 24 distinct optimum design schemes for RA and 48 distinct optimal design schemes for GA. Analyses were conducted to evaluate the performance of energy, water, and carbon on a case-by-case basis.

The optimization results for the on-farm stage showed significant improvements. Energy consumption per joule of food production in the RA system decreased to 6.41 × 10⁻⁶ MJ (Figs. 5 (c)–(e), mean value, the same goes after). Water consumption decreased to 6.93 × 10⁻⁶ L, whereas carbon emissions decreased to 3.57 × 10⁻⁷ kg. Similarly, through the optimization of GA system, the energy consumption, water consumption and carbon emissions per joule of food production in GA were reduced to 7.12 × 10⁻⁶ MJ, 6.93 × 10⁻⁶ L, and 6.02 × 10⁻⁷ kg, respectively. During the off-farm stage, GA had the impacts of energy, water, and carbon decline by 80.4%, 61.4%, and 54.0%, respectively. Given that the environmental performance of both aquaponics systems during the off-farm stage surpassed that of TG, it can be concluded that the optimized aquaponics system has superior performance. The optimized aquaponics system shows significant potential for energy improvement during the on-farm stage, and the average energy consumption is 1.46–1.48 times greater than that of TG. This highlights the continued competitiveness of TG in terms of energy consumption in the future. After optimization, the water usage of the two aquaponics systems safely decreased compared with that of the original design. The reduction ratio for RA ranges from 0 to 27%, while for GA, it ranges from 0 to 35%. The potential for energy and carbon reduction is significant, particularly due to the ability of GA systems to decrease energy consumption by up to 80% and lower carbon emissions by 85% (Figs. 5 (f) and (g)).

3.4. Scenario analysis

When considering urban PA (including RA, GA, and TG in this research), the three future scenarios (Scenarios 1, 2, and 3) demonstrate resilience improvement in terms of the local food supply. This is evident via a self-sufficiency rate of 24.8%–26.2% provided by PA, which is higher than the 9.88% self-sufficiency rate seen in the BAU scenario (Fig. 6). Simultaneously, Scenarios 1, 2, and 3, which are characterized by varied food farming systems can offer supplementary fruits and marine products, enhancing the diversity of local food options and the dietary consumption of local dwellers. More precisely, Scenario 1 has the highest overall vegetable output, amounting to 2.88 × 10⁶ t, and it can also provide an extra 2.14 × 10⁴ t of strawberries, which is comparable to the 6% of Beijing’s fruit production in 2022. The total vegetable yield of Scenario 2 (2.73 × 10⁶ t) is 5% lower than that of Scenario 1, but it achieves the most striking increase in fish production, with an extra 1.17 × 10⁶ t. Vegetable culture is often seen as the primary focus in aquaponics systems because it tends to provide more revenue than fish production [11]. This research examines fish as a by-product, which may also serve as a substantial supplementary revenue source for sellers. The results indicate that the ratio of vegetable to fish output is approximately 2.3 for Scenario 2 and 4.7 for Scenario 3, whereas it is 102.7 for Scenario 1. Urban decision-makers’ desire for fruit or fish harvest may determine the optimal development scenario, when assuming equivalent vegetable yields.

From an economic perspective, Scenario 2 has the highest economic cost (6.79 × 10¹⁰ CNY) and highest economic income (1.21 × 10¹¹ CNY), while the lowest economic cost and income both come from Scenario 1, which has the largest percentage of TG, amounting to less than half of Scenario 2 (Table 2). By considering the economic cost and income, the profits of Scenarios 2 and 3 are 5.26 × 10¹⁰ and 2.85 × 10¹⁰ CNY, respectively, which are much greater than the profits of Scenario 1, highlighting the substantial economic advantages that aquaponics systems provide to aquaponics operators. From the standpoint of resource and environmental impacts, in the on-farm stage, Scenario 1 drives the most water consumption (Fig. 7(a)), while the highest energy consumption and carbon emissions are sourced from Scenario 2 (Fig. 7(a)). Although Scenario 1 has the largest yield, which is 2.65-fold greater than that of the BAU scenario, the energy and water consumption and the carbon emissions of Scenario 1 are only 1.32–1.39 times greater than those of the BAU scenario. This finding demonstrates that, compared with BAU, Scenario 1 results in superior production efficiency while also achieving a more ecologically sustainable performance. The disparity in vegetable yield between Scenarios 1 and 3 is not statistically significant. However, the water usage of Scenarios 2 and 3 is only 33%–67% of that of Scenario 1. This is mostly due to the notable water recycling advantages of aquaponics over TG.

Despite the initial underperformance of the aquaponics system in terms of carbon and energy, implementing the improved aquaponics system results in significant carbon mitigation for Scenarios 2 and 3. Specifically, the carbon emissions of Scenarios 2 and 3 are merely 56%–78% of those of Scenario 1. Meanwhile, the energy consumption of Scenarios 2 and 3 remains high, ranging from 1.18–1.37 times that of Scenario 1. This is mostly because the average energy and carbon performance of the optimized aquaponics system are still superior to those of TG. If ideal performance (integrated optimization of fish food, electricity, and infrastructure materials) is achieved, then the energy consumption of Scenarios 2 and 3 can reach 58%–79% of that of Scenario 1. Nevertheless, carbon emissions had a more favorable outcome, ranging from 30%–65%, whereas water consumption showed a similar improvement, ranging from 28%–64%. Put simply, if optimal aquaponics can be expanded to include whole cities, then cities can achieve greater food self-sufficiency while minimizing their environmental impacts.

In the off-farm stage (Fig. 7(b)), the aquaponics scenario demonstrates significant environmental friendliness, with energy consumption, water consumption, and carbon emissions in Scenarios 2 and 3 being 0.7%–50%, 16%–58%, and 5%–53% of those in Scenario 1, respectively. The BAU scenario requires the greatest upstream food supply (81.9%) Fig. 6(a) for the city’s vegetable self-sufficiency. Since Scenarios 1, 2, and 3 increase the local self-sufficiency of vegetables by 18%–20%, large amounts of energy, water, and carbon footprints in the upstream food supply can be avoided. Notably, when all potential areas for urban RA and GA systems are implemented, developing 154.26 km² of aquaponics areas at the city scale can increase local vegetable self-sufficiency by 15%. It further results in the avoidance of 82% of the energy, water, and carbon footprints during the upstream food supply chain beyond cities compared with BAU scenario, reaching values of 2.20 × 10²⁷ MJ, 2.07 × 10²⁸ L, and 7.69 × 10²⁵ kg, respectively (Table 3). This finding shows that based on the multiple optimal designs, urban aquaponics not only presents well in the local aspects of life cycle impact reduction and potential economic benefits but also plays a vital role in the transboundary environment footprint savings in the processes of upstream food supply.

4. Discussion

Given that sustainable production is a key objective of the United Nations’ Sustainable Development Goals, it is imperative for agriculture operators and consumers to collaborate to minimize the environmental impacts of the food supply process [29]. The simplification of agricultural systems continues to come at the expense of more diversified agriculture, with planetary boundaries being crossed because of the overuse of chemical inputs, greenhouse gas emissions, biodiversity loss, and increased water consumption [30]. To address these challenges, a new paradigm for agricultural systems is needed that focuses on providing food security and nutrition while minimizing the negative environmental, health, and social impacts [31], [32]. Aquaponics is an emerging technique that addresses the increasing global food demand by efficiently recycling water and nutrients, resulting in higher productivity but less water and nutrient waste than traditional agriculture. Additionally, aquaponics enhances resilience by creating decentralized, urban-based systems that provide a reliable and local source of fresh produce. This reduces dependency on long supply chains, which are vulnerable to disruptions [33], [23]. By integrating aquaponics into food production, we can achieve both environmental and economic benefits, which makes it a promising solution for future food systems that are both sustainable and resilient.

This study not only establishes a flexible and general methodology and framework to capture the FWE nexus flows of both RA and GA systems at the urban scale but also provides evidence-based case studies and policy insights into how to scale up sustainable UA technologies. Here, we establish a generalizable methodology and framework for assessing the impacts of aquaponics on the FWE nexus at the city scale. The framework includes opportunity maps of urban aquaponics, life cycle assessment through on- and off-farm stages, sensitivity analysis and optimization, and tradeoff analysis associated with upscaling aquaponics across the whole city. Given that each study varies and has unique difficulties, researchers can accordingly modify the life cycle boundaries and lists of study items based on their objectives. The proposed framework is anticipated to serve as a decision-making tool for urban stakeholders, enabling them to modify agricultural production techniques to reduce environmental impacts, particularly in terms of energy consumption and feed formulations.

The shifting consumption habits in emerging nations highlight the significance of the nutritional composition of food sources, particularly meat, fruits, and vegetables. Aquaponics is a viable alternative for sustainable agriculture since it enables greater production quality and agricultural output. Nevertheless, aquaponics requires electrical input, which strains the environment. The primary objective of aquaponics is to minimize energy consumption and mitigate the environmental effects associated with energy generation. This may be achieved via the implementation of two distinct means: utilizing sustainable energy sources and minimizing water extraction for pumping purposes [34]. Shifting from coal-based power to renewables such as wind or photovoltaic energy could be a feasible solution to mitigate the impacts of electricity from thermal power. To minimize energy consumption, it is also possible to achieve energy savings by decreasing the quantity of pumps used and optimizing the water flow. Furthermore, increasing the number of cultivation beds allocated to each pump is suggested. Optimizing the equilibrium between these two elements may effectively restrict energy use, thereby enhancing system profitability. In addition to electricity usage, the use of fish feed and infrastructure materials poses significant risks to the sustainability of aquaponics products. The impacts of aquaponics on the urban environment in terms of energy, water, and carbon can be mitigated by replacing materials, fish food components, and electricity sources and increasing recycling practices. The combination of various strategies can reduce by 85% energy consumption and carbon emissions, which are the most impactful aspects of an aquaponics system during the on-farm stage. By avoiding the need for supermarkets and consumers to go out to buy food, the direct home delivery of food adopted by aquaponics can greatly prevent the negative environmental impacts of this stage. The whole life cycle cost–benefit analyses also reveal the positive economic returns of aquaponics.

Furthermore, including vegetable production in aquaponics systems will result in further water conservation. The significance of these water conservation efforts should not be diminished by the substantial energy requirements. Agriculture confronts the contradictory demands of significantly decreasing water use while simultaneously augmenting output to fulfill the dietary requirements of a growing population. According to the analytical findings of this research, aquaponics systems are considered a valuable method for enhancing water efficiency in food production. Nevertheless, we also note the overwhelming contribution of irrigation water to the life cycle water consumption in aquaponics systems during the operational period, which was observed to be more severe in TG. We emphasize the need to enhance irrigation systems to reduce water intensity and the transition of pumping systems to electric pumping to reduce energy and CO₂ emissions while utilizing low-carbon electricity [35].

One of the worldwide concerns related to agricultural production is nutrient release [36]. TG and open-air agriculture tend to have enormous pesticide use requirements, while aquaponics can cut or eliminate the use of pesticides [37], thus mitigating environmental impacts, including decreased carbon emissions and nitrogen loss. This is because the nutrients produced by fish in aquaponics systems are recycled and used as fertilizers for plant growth, thus offsetting the production of similar amounts of synthetic nitrogen and phosphorus fertilizers; therefore, no additional nitrogen or phosphorus is added [38]. Otherwise, the release of nitrogen and phosphorus would likely contribute 30% of the eutrophication potential [39]. Indeed, there may also be interlinkages and externalities between aquaponics and TG in terms of pesticide use. A recent study revealed that at lower rates of organic farming, overall pesticide use increases due to positive spillovers from organic to traditional farmland (possibly an increase in pests), whereas at higher rates of organic farming, pesticide use decreases due to direct effects and spillover effects [37]. This highlights the important influence of the spatial configuration of aquaponics versus TG on pesticide use and pest management strategies and highlights the potential for reducing pesticide use and improving agricultural sustainability through rational planning and the promotion of spatial aggregation in organic agriculture.

This research identifies potential areas for urban RA. Meanwhile, the building height and load-bearing structure are excluded when the opportunity map for urban RA is identified. Future mapping of opportunities for RA should consider not only the key constraints of building load-bearing structures [[40], [41]] and height [42] but also a wider array of structural, environmental, and logistical factors. For example, rooftop orientation and solar exposure play critical roles in determining the viability of aquaponics, as sunlight availability directly influences plant growth and the energy requirements for supplemental lighting. Wind loads at different building heights may also affect the stability and durability of rooftop systems, necessitating careful planning and the potential use of windbreaks or protective barriers. Moreover, geographic information systems (GIS) sampling methods can be extended to incorporate additional spatial variables such as proximity to water sources, local climate conditions, and building density, which all contribute to the feasibility of RA at larger scales. In terms of energy consumption, the pumping requirements for water irrigation across different building heights must be factored into a comprehensive life cycle assessment framework. This framework should not only account for the direct energy use involved in pumping water to various floors but also consider how factors such as building height, pump efficiency, water source location, and irrigation demands affect overall system performance. By incorporating these variables, the assessment can offer a more nuanced understanding of the energy tradeoffs involved in RA systems, especially compared with ground-level alternatives.

Here, we recognize the critical role of urban aquaponics in the future of agricultural transition pathways with increasing local food self-sufficiency and biodiversity, but it is also important to acknowledge that many farmers are working “against the odds”. Structural factors are often major barriers to diversification, including high land rents, the predominance of short-term leases, stringent food safety regulations, trade agreements that exacerbate the concentration of businesses in the global food system, and other supply chain pressures [30]. Transitions to aquaponics systems often require financial support because of potential initial yield declines or implementation costs [43]. Current policies are often limited to simplified traditional agriculture rather than achieving a lasting transition to aquaponics, and investments are needed to develop appropriate seeds, crop combinations, rotations, and equipment to improve the profitability of site-scale agriculture [44]. Effective policies to encourage the adoption of diversification strategies and practices may vary by cropping system and region, including incentives, regulations, and integrated approaches [45]. The use of incentives can be seen in the European Union, where farmers are financially compensated on a per-region basis for some diversification practices (e.g., non-crop cultivation).

When developing aquaponics systems in different regions, various factors, such as environmental, technical, economic, social, and policy considerations, must be considered. First, climate conditions, sunlight availability, and water resources directly impact system design, thus requiring tailored solutions. Tropical and subtropical regions, because of their warm climates and abundant resources, have relatively simple system operations and lower development difficulties. The development difficulty in temperate and arid regions is relatively high, and these regions face challenges such as low winter temperatures and scarce water resources. Technically, it is essential to select the appropriate system design based on regional characteristics and incorporate renewable energy sources such as solar power to minimize energy consumption. Economically, assessing initial investment, operational costs, and local market demand is crucial to ensure the system’s financial viability. Different planting types depend on market value and demand, and future aquaponic systems can be developed into multiple vegetable planting types to promote an aquaponic system that synergistically cultivates various vegetable crop types to more comprehensively meet the nutritional needs of residents’ daily lives. Social aspects, such as technical support, education, and local dietary preferences, should also be considered, with appropriate fish and plant species selected to meet local needs. Finally, complying with environmental regulations and land use policies and leveraging government subsidies and support programs are key to achieving sustainable development.

Our empirical cases were conducted at a site-scale facility, and neither RA nor GA were operated under optimized conditions. The optimal system solutions should consequently be put into reality for improvement, including infrastructure material, fish food replacement, and renewable energy input. Since we compared system-specific designs rather than standard production protocols, the results may not be automatically generalized to existing aquaponics systems but could be used to support the general concept of their environmental benefits. Public preferences are highly important when promoting a new paradigm of UA [46]. Previous studies have shown that consumers’ willingness to pay for the benefits of aquaponics is limited [47], [48], as is the level of adoption of aquaponics by growers, largely due to low profitability [11]. The current study suggests that the environmental benefits of aquaponics have considerable economic value that is not present when alternative production systems are chosen solely for economic gain. Social interventions in support of aquatic production may mitigate this existing market failure. In the future, more field surveys on the extent of realistic policy support for aquaponics should also be conducted, as this is an important cornerstone of sustainable urban agricultural transformation based on the FWE nexus [49]. Case studies should also be carried out in more cities to enhance the understanding of the heterogeneity of the FWE nexus in urban aquaponics. Aquaponics evaluation could benefit from further research into life cycle assessment, which could lead to a better understanding of the environment–nutrition interface through the creation of context-specific databases [18]. These databases would help stakeholders, including consumers and agri-food policy-makers, obtain a better picture of the foods available to them in their “foodsheds” by considering factors such as local production conditions and the number of food types that have a similar nutritional value.

5. Conclusions

Urban aquaponics presents a promising solution to enhance urban food security and resilience, by integrating circular aquaculture with hydroponics to produce fish and vegetables sustainably. This research assesses the geospatial potential, environmental and economic performances of large-scale aquaponics in Beijing. It is revealed that by integrating renewable energy, optimizing fish feed, and improving material choices, urban aquaponics can effectively reduce energy and water consumption, as well as carbon emissions. Scaling urban aquaponics across Beijing could increase local vegetable self-sufficiency and abate the environmental impacts through the upstream food supply chain. This study underscores the potential for urban aquaponics to transform cities into self-sufficient “edible cities” with reduced environmental footprints. By offering a generalizable method framework, we lay a foundation for assessing and optimizing future agricultural modes, and provide insights for urban stakeholders aiming to enhance the agricultural sustainability. Furthermore, the findings encourage urban consumers to consider the environmental benefits of locally grown produce from aquaponics, fostering greater awareness and demand for environmentally responsible products.

CRediT authorship contribution statement

Qiuling Yuan: Writing – original draft, Visualization, Methodology, Conceptualization. Fanxin Meng: Writing – review & editing, Data curation, Supervision, Funding acquisition, Conceptualization. Yingxuan Liu: Visualization, Methodology, Data curation. Jose A. Puppim de Oliveira: Writing – review & editing, Data curation. Lixiao Zhang: Writing – review & editing, Funding acquisition. Wenting Cai: Data curation. Zhifeng Yang: Writing – review & editing, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (72174028 and 52225902), and the Strategic Research and Consulting Project of Chinese Academy of Engineering (2024-XZ-47). The authors are thankful to the Beijing Futian Agricultural Technology Co., Ltd. for supporting our field investigation and especially to Xu Cheng for the data contribution and detailed interview feedback.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.eng.2025.01.021.

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Seto KC, Ramankutty N. Hidden linkages between urbanization and food systems. Science 2016; 352(6288):943-945.

[2]	Lal R. Home gardening and urban agriculture for advancing food and nutritional security in response to the COVID-19 pandemic. Food Secur 2020; 12(4):871-876.

[3]	Hawes JK, Goldstein BP, Newell JP, Dorr E, Caputo S, Fox-Kämper R, et al. Comparing the carbon footprints of urban and conventional agriculture. Nat Cities 2024; 1(2):164-173.

[4]	Russo A, Escobedo FJ, Cirella GT, Zerbe S. Edible green infrastructure: an approach and review of provisioning ecosystem services and disservices in urban environments. Agric Ecosyst Environ 2017; 242:53-66.

[5]	Sanjuan-Delmás D, Llorach-Massana P, Nadal A, Ercilla-Montserrat M, Muñoz P, Montero JI, et al. Environmental assessment of an integrated rooftop greenhouse for food production in cities. J Clean Prod 2018; 177:326-337.

[6]	Bellezoni RA, Meng F, He P, Seto KC. Understanding and conceptualizing how urban green and blue infrastructure affects the food, water, and energy nexus: a synthesis of the literature. J Clean Prod 2021; 289:125825.

[7]	Caputo S, Schoen V, Specht K, Grard B, Blythe C, Cohen N, et al. Applying the food–energy–water nexus approach to urban agriculture: from FEW to FEWP (food–energy–water–people). Urban For Urban Green 2021; 58:126934.

[8]	Meng F, Yuan Q, Bellezoni RA, de JAP Oliveira, Cristiano S, Shah AM, et al. Quantification of the food–water–energy nexus in urban green and blue infrastructure: a synthesis of the literature. Resour Conserv Recy 2023; 188:106658.

[9]	Meng F, Yuan Q, Bellezoni RA, de JAP Oliveira, Hu Y, Jing R, et al. The food–water–energy nexus and green roofs in Sao Jose dos Campos, Brazil, and Johannesburg, South Africa. NPJ Urban Sustain 2023; 3(1):12.

[10]	Ghamkhar R, Hartleb C, Rabas Z, Hicks A. Evaluation of environmental and economic implications of a cold-weather aquaponic food production system using life cycle assessment and economic analysis. J Ind Ecol 2022; 26(3):862-874.

[11]	Love DC, Fry JP, Li X, Hill ES, Genello L, Semmens K, et al. Commercial aquaponics production and profitability: findings from an international survey. Aquaculture 2015; 435:67-74.

[12]	Rakocy JE. Aquaponics—integrating fish and plant culture. J.H. Tidwell (Ed.), Aquaculture production systems, John Wiley & Sons, Hoboken 2012; 344-386.

[13]	Goddek S, Joyce A, Kotzen B, Burnell GM.. In: Aquaponics food production systems: combined aquaculture and hydroponic production technologies for the future. Berlin: Springer Nature; 2019. p. 619.

[14]	Meyfroidt P, de A Bremond, Ryan CM, Archer E, Aspinall R, Chhabra A, et al. Ten facts about land systems for sustainability. Proc Natl Acad Sci USA 2022; 119(7):e2109217118.

[15]	Seto KC, Reenberg A, Boone CG, Fragkias M, Haase D, Langanke T, et al. Urban land teleconnections and sustainability. Proc Natl Acad Sci USA 2012; 109(20):7687-7692.

[16]	Lobell DB, Villoria NB. Reduced benefits of climate–smart agricultural policies from land-use spillovers. Nat Sustain 2023; 6(8):941-948.

[17]	Baganz GF, Timpe A, Baganz D, Staaks G, Hunger B, Kloas W, et al. City or hinterland—site potentials for upscaled aquaponics in a Berlin case study. NPJ Urban Sustain 2022; 2(1):29.

[18]	Katz-Rosene R, Ortenzi F, McAuliffe GA, Beal T. Levelling foods for priority micronutrient value can provide more meaningful environmental footprint comparisons. Commun Earth Environ 2023; 4(1):287.

[19]	Yuan Q, Mi S, Meng F, Hou J, Sun Y, Li H, et al. An emergy analysis of environmental sustainability in urban agriculture: evidence from protected agriculture in Beijing, China. Front Sustain Food Syst 2023; 7:1288136.

[20]	Hu Y, Zheng J, Kong X, Sun J, Li Y. Carbon footprint and economic efficiency of urban agriculture in Beijing—a comparative case study of conventional and home-delivery agriculture. J Clean Prod 2019; 234:615-625.

[21]	Newell JP, Ramaswami A. Urban food–energy–water systems: past, current, and future research trajectories. Environ Res Lett 2020; 15(5):50201.

[22]	Saha M, Eckelman MJ. Growing fresh fruits and vegetables in an urban landscape: a geospatial assessment of ground level and rooftop urban agriculture potential in Boston, USA. Landsc Urban Plan 2017; 165:130-141.

[23]	Zhang L, Huang L, Li T, Wang T, Yang X, Yang Q. The skyscraper crop factory: a potential crop-production system to meet rising urban food demand. Engineering 2023; 31:70-75.

[24]	Mao R, Bao Y, Huang Z, Liu Q, Liu G. High-resolution mapping of the urban built environment stocks in Beijing. Environ Sci Technol 2020; 54(9):5345-5355.

[25]	Shi M, Lu X, Jiang H, Mu Q, Chen S, Fleming RM, et al. Opportunity of rooftop solar photovoltaic as a cost-effective and environment-friendly power source in megacities. iScience 2022; 25(9):104890.

[26]	Beijing Municipal Bureau of Statistics. Beijing Statistical Yearbook 2023.Beijing: China Statistics Press; 2023. Chinese.

[27]	Ghamkhar R, Hartleb C, Wu F, Hicks A. Life cycle assessment of a cold weather aquaponic food production system. J Clean Prod 2020; 244:118767.

[28]	Boxman SE, Kruglick A, McCarthy B, Brennan NP, Nystrom M, Ergas SJ, et al. Performance evaluation of a commercial land-based integrated multi-trophic aquaculture system using constructed wetlands and geotextile bags for solids treatment. Aquacult Eng 2015; 69:23-36.

[29]	Carriquiry M, Dumortier J, Elobeid A. Trade scenarios compensating for halted wheat and maize exports from Russia and Ukraine increase carbon emissions without easing food insecurity. Nat Food 2022; 3(10):847-850.

[30]	Rasmussen LV, Grass I, Mehrabi Z, Smith OM, Bezner-Kerr R, Blesh J, et al. Joint environmental and social benefits from diversified agriculture. Science 2024; 384(6691):87-93.

[31]	Goldstein B, Hauschild M, Fernández J, Birkved M. Testing the environmental performance of urban agriculture as a food supply in northern climates. J Clean Prod 2016; 135:984-994.

[32]	Springmann M, Clark M, Mason-D' D Croz, Wiebe K, Bodirsky BL, Lassaletta L, et al. Options for keeping the food system within environmental limits. Nature 2018; 562(7728):519-525.

[33]	Weidner T, Yang A, Forster F, Hamm MW. Regional conditions shape the food–energy–land nexus of low-carbon indoor farming. Nat Food 2022; 3(3):206-216.

[34]	Mohareb E, Heller M, Novak P, Goldstein B, Fonoll X, Raskin L. Considerations for reducing food system energy demand while scaling up urban agriculture. Environ Res Lett 2017; 12(12):125004.

[35]	Qin J, Duan W, Zou S, Chen Y, Huang W, Rosa L. Global energy use and carbon emissions from irrigated agriculture. Nat Commun 2024; 15(1):3084.

[36]	Joyce A, Goddek S, Kotzen B, Wuertz S. Aquaponics: closing the cycle on limited water, land and nutrient resources. In: Goddek S, Joyce A, Kotzen B, Burnell GM, editors. Aquaponics food production systems. Berlin: Springer; 2019.

[37]	Larsen AE, Noack F, Powers LC. Spillover effects of organic agriculture on pesticide use on nearby fields. Science 2024; 383(6689):eadf2572.

[38]	Bernstein S. Aquaponic gardening: a step-by-step guide to raising vegetables and fish together. Gabriola Island: New Society Publishers; 2011.

[39]	Bordignon F, Sturaro E, Trocino A, Birolo M, Xiccato G, Berton M. Comparative life cycle assessment of rainbow trout (Oncorhynchus mykiss) farming at two stocking densities in a low-tech aquaponic system. Aquaculture 2022; 556:738264.

[40]	Jing R, Liu J, Zhang H, Zhong F, Liu Y, Lin J. Unlock the hidden potential of urban rooftop agrivoltaics energy–food-nexus. Energy 2022; 256:124626.

[41]	Yuan Q, Meng F, Li W, Lin J, de JAP Oliveira, Yang Z. Tradeoff optimization of urban roof systems oriented to food–water–energy nexus. Appl Energ 2025; 380:124987.

[42]	Chen Y, Liu Y, Slootweg M, Hu M, Tukker A, Chen W. Unlocking rooftop potential for sustainable cities: a systematic review. Front Eng Manag 2024; 1-15.

[43]	Zhu Z, Chan FKS, Li G, Xu M, Feng M, Zhu YG. Implementing urban agriculture as nature-based solutions in China: challenges and global lessons. Soil Environ Health 2024; 2(1):100063.

[44]	An Z, Yang Y, Yang X, Ma W, Jiang W, Li Y, et al. Promoting sustainable smallholder farming via multistakeholder collaboration. Proc Natl Acad Sci USA 2024; 121(21):e2319519121.

[45]	Grabs J, Carodenuto S, Jespersen K, Adams MA, Camacho MA, Celi G, et al. The role of midstream actors in advancing the sustainability of agri-food supply chains. Nat Sustain 2024; 7:527-535.

[46]	Mann S, Kaiser A. Why is agricultural policy not more environmentally ambitious? Comparing failed attempts in Switzerland. Resour Environ Sust 2023; 11:100096.

[47]	Greenfeld A, Becker N, Bornman JF, Angel DL. Identifying knowledge levels of aquaponics adopters. Environ Sci Pollut Res Int 2020; 27:4536-4540.

[48]	Short G, Yue C, Abbey M, Anderson N, Phelps N, Venturelli P, et al. Consumer preferences for aquaponic produce: implications from an experimental auction. Agribusiness 2018; 34(4):742-755.

[49]	Zhuang J, Gill T, Löffler FE, Jin M, Sayler GS. Can food–energy–water nexus research keep pace with agricultural innovation?. Engineering 2023; 23:24-28.

RIGHTS & PERMISSIONS

THE AUTHOR