The use of complex adaptive system’s emulation and principles in planning and managing a biophilic systems transition in Singapore

Abstract

Cities have been described as complex adaptive systems (CAS). A key aspect of both natural and social CAS is that they are scaler and hierarchical. The research reported in this article looked at the purposeful adoption of CAS principles within the governance structures, socioecological and sociotechnical subsystems of the city state of Singapore in order to deliver urban ecological services. The major objective of the research was to evaluate the degree to which complex adaptive systems restoration, rehabilitation, and emulation and/or principles are holistically being applied across sectors, departments, and agencies to deliver urban ecological services that reach normative goals. The research first analysed the Singapore government’s intersectoral and sectoral ‘Big P’ policy documents. It then analysed whether policies within these are subsequently being incorporated into the ‘small p’ regulatory standards and guidelines as well as projects of relevant departments. The research found the most ‘Big P’ policy documents and to a lesser extent ‘small p’ standards and projects have purposely adopted nature-based solutions in order to deliver intersectoral urban ecological services. However, adoption is often voluntary, as a result, it is largely progressing via pilot schemes and ad hoc projects in places of least resistance. As such, it is not yet being undertaken as a holistic coordinated city-wide transition.

Introduction

Cities have been described as complex adaptive systems (CAS) (Meerow et al. 2016, Bush and Doyon 2019). In nature, complex adaptive systems are the structures behind biological evolution that have seen life evolve from the primordial swamp to the complex and diverse array of species and subspecies we see in nature today (Dawkins 2009). The dynamic of CAS is that they endogenously produce collective intelligence, strength and abilities arising from the agglomeration and interconnection of often fragile agents, making simple decisions, bound by simple rules (Dawkins 2009, Mitchell 2009, Page 2010).

Understanding city systems as complex and/or using complexity principles in urban system’s organisation, analysis and management has been applied to areas as diverse as governance (Booher and Innes 2002, McGreevy et al. 2020), transport and logistic systems (Reggiani et al. 2015), economic resilience (Martin and Sunley 2015), expediting sustainability transitions (Geels 2005), designing activity centres (Childs 2006, McGreevy and Wilson 2016), rewilding (Carver et al. 2021), stormwater management (Finewood et al. 2019) and urban based permaculture (Ferguson and Lovell 2014).

This article reports on research looking at complexity theory as it has been applied in governance systems and the socioecological systems of the city of Singapore in order to deliver beneficial social, economic, and environmental urban ecological services (UES). This includes systems emulation which draws upon principles, species, and relationships found in natural CAS. The natural complex adaptive systems that have the most similarities with urban systems are ecosystems. Both contain a rich diversity of biotic agents (flora and fauna in the case of ecosystems; humans, institutions, governments, communities and businesses, as well as flora and fauna in the case of cities) along with abiotic elements (terrain, water bodies and topography in the case of ecosystems; buildings and infrastructure as well as terrain, water bodies and topography in the case of cities). These in combination create systems of rich and exquisite interconnected diversity in both (Collins et al. 2000). The key to the agglomerated strength of both systems is a rich diversity of individual agents bounded and connected via networks of relationships that produce catalysts, information loops, synergies, and emergent phenomena (Johnson 2002, Mitchell 2009).

This research uses an urban ecology approach to analyse the socioecological systems of the city. Urban ecology is a scholarship that understands cities as socioecological systems consisting of the hard surfaces of buildings and infrastructure, and the soft surfaces of public and private spaces that lie between them (Breuste et al. 2013). Within the greater socioecological system of the city are socioecological subsystems which include nature reserves, parks, sites inappropriate or problematic for development, cemeteries, buffers, private lots, verges etc. These can individually or in concert give rise to complex human and non-human activity and beneficial UES (Breuste et al. 2013, Larondelle and Haase 2013).

The UES benefits of socioecological systems within the city have resulted in efforts to preserve, expand, and retrofit them. These efforts have included the full or partial reintroduction, rehabilitation, and/or restoration of lost or degraded natural ecosystems (Garcia 2017, Pedersen Zari 2017). It has also included bioengineering that uses natural biotic and abiotic elements in conjunction with synthetic materials and devices to emulate or mimic phenomena found in natural complex adaptive ecosystems, such as forests or reefs, in order to deliver urban ecological services such as hazard control, pest control, and stormwater management (Garcia 2017). Ecosystem emulation or mimicry can be of the form, materials, construction processes, principles, or functions such as interspecies relationships, synergies, and emergence (Pedersen Zari 2017). This includes bio-sensitive urban design which incorporates ecological knowledge into urban planning and development to produce UES (Garrard et al. 2018). There is also the application of ecosystem and bioengineering to the specific task of stormwater management. These projects have been given names and acronyms, such as Low Impact Development (LID), Low Impact Urban Design and Development (LIUDD), Stormwater Best Management Practices (BMPs), Sustainable Urban Drainage Systems (SUDS), Sponge Cities, Nature-Based Solutions (NBS), and Water Sensitive Urban Design (WSUD) (Fletcher et al. 2015). For practicality, this article will only use the acronym WSUD.

Singapore is looking at transitioning many of its urban systems with nature based ecological solutions to help solve urban problems and produce desirable UES, including in stormwater management. International studies have shown that government expedited transitions such as these can encounter resistance, path dependency, reinterpretation, compromise, misunderstanding or unintended consequences and require intersectoral coordination (Sabatier 1986; Hill and Hupe 2002, Nilsen et al. 2013). Scholars have identified the gaps between policy intent and policy implementation in complex policy arenas as significant obstacles to coordinated holistic transitions (Hill and Hupe 2002, Nilsen et al. 2013). These gaps can be found in both how well and how broadly policies are applied via governance systems and organizational arrangements (Spillane et al. 2002, Hill 2005). Within these systems, gaps or contradictions can be found between the ‘Big P’ policies of central governments and the ‘small p’ rules (guidelines, regulations, standards etc.) that determine day-to-day decision making (Nilsen et al. 2013) as well as the capacity, resources, skill, knowledge values, norms, and/or practices of agents charged with implementing them (Sabatier 1986, Spillane et al. 2002). Furthermore, policies intended to expedite change toward normative goals in complex policy arenas can encounter resistance, path dependency, reinterpretation, compromise, misunderstanding or unintended consequences at the implementation stage (Sabatier 1986, Hill and Hupe 2002, Nilsen et al. 2013).

The translation of this research is guided by transition management theory. Transition management theory is most associated with expediting the transition of energy systems towards the goal of environmental sustainability (Köhler et al. 2019). It recognises the plurality of actors large and small within and outside government involved in transition management and advocates the use of collaborative planning and joined-up governance structures which incorporate heterarchical system structures, inter-sectoral cooperation and coordination, and reflexive frameworks as a means of guiding transitions towards normative goals (Reff Pedersen et al. 2010, Potts et al. 2014, Jørgensen et al. 2017). Collaborative planning and joined-up governance approaches also reflect principles from complexity theories (Chettiparamb 2016, Innes and Booher 1999).

Singapore is an interesting case study in both urban ecology and policy implementation and governance. It is a city-state of just 720 km² with a growing population approaching 5.9 million. It has the highest per capita GDP in the tropics and has some of the best R&D infrastructure and expertise in the world (NRF 2020). Therefore, it is in a strong position to lead advances in socioecological and sociotechnical transition management which can then inform cities with similar climates globally.

Singapore has several features that make it unique, some of which make it a strong candidate for creating joined-up governance or collaborative planning structures. It is a city-state, therefore, many of the basic functions of a nation such as defence, supplying food, water, and energy, as well as waste processing and disposal must be undertaken within the confines of the city. It is routinely reiterated that Singapore’s most scarce and therefore valuable resource is land, and it cannot be wasted and needs to be allocated with maximum efficiency (CLC 2016, Tan 2019). A notable aspect of Singapore is the high level of government land ownership. Approximately 90% of land in Singapore is owned by the government and then used directly for public uses or leased to private users (Good 2019). Figure 1 shows the distribution of land uses across the country. In addition, there is a single level of government that has been dominated by a single political party since independence in 1965 which routinely and directly intervenes in all sectors of the economy and society (Han 2017, Woo 2018). This sets its governance systems apart from many other nations and cities (Ufen 2015).

The research analyses the adoption of the ‘Big P’ policies outlined in the Singapore Green Plan (2020a) into multisectoral ‘Big P’ and ‘small p’ policies and projects. The Singapore Green Plan is the nation’s overarching plan for a long-term sustainability transition. It includes policies focussed upon both climate change mitigation and adaptation. The long-term goal is a closed-loop, resilient, self-reliant, and zero greenhouse gas emitting society and economy. A key component of this is the incorporation of nature-based solutions as a means of overcoming climate related threats. The Green Plan contains 10 targets in multiple sectors and includes five ‘key pillars’ to achieve these targets:

• City in Nature

• Energy Reset

• Sustainable Living

• Green Economy

• Resilient Future

This research looks at the extent to which Singapore’s transitions towards nature-based solutions advocated in the Green Plan under the key pillar of City in Nature are percolating down into the lower order policies and projects overseen by government departments and agencies and whether the intersectoral benefits of UES are considered. The aim is to identify a holistic coordinated approach reflective of complexity-based governance. Another aim is to evaluate the degree to which nature-based solutions and intersectoral UES provision are being hampered by the obstacles observed in transition planning scholarship. The research takes a dual approach to achieving this aim. The first is an analysis of the ‘Big P’ policy documents from relevant government agencies. The second is an analysis of whether these policies are being transmitted holistically into coordinated outcomes. This includes their inclusion in ‘small p’ regulatory standards and guidelines and recent urban development projects.

The article is in four sections. This includes the introduction. The background section looks at areas where complexity theories are being used to deliver urban ecological services (UES) and the requirements for transitions toward their holistic adoption. This section also provides greater information about Singapore. Section three outlines the methods used in the research. Section four presents the findings from document analysis and observation of projects. Section five further analyses and discusses the findings.

Background

A principle of natural complex adaptive ecosystems, which has application to city systems, is that systems are intertwined and connected and can provide sustenance and benefits to other systems as well as emergent phenomena (Page 2010, Johnson 2012). Similarly, socioecological systems can provide direct and emergent urban ecological services. These include UES such as habitats for wildlife, soil formation, bio-diversity refuges, pollination, microclimate moderation, as well as pollution mitigation, noise abatement, and reductions from hazards such as fire, flood, and landslides (Gómez-Baggethun and Barton 2013, Larondelle and Haase 2013). Biodiverse greenspaces also provide humans with opportunities for exercise, communion with nature, socialising and relaxation with major benefits for population health and wellbeing (Hooper et al. 2015, Astell-Burt and Feng 2019). Furthermore, well maintained, and regular trees, shrubs and smaller plants along roads provide canopy, shelter and shade, barriers between pedestrians and traffic, human scale, signs of care, interestingness, and sensual stimulation. In doing so, they help mitigate heat island effects, calm traffic, and make the public realm safer, more attractive, pleasant, and healthier (Ewing and Handy 2009, Breuste et al. 2013, Mahdjoubi and Spencer 2015). However, green infrastructure in the city can have unintended side effects such as becoming a habitat for disease vectors (Frantzeskaki 2019).

Internationally, there have been numerous nature-based projects that protect, rehabilitate, and (re)introduce ecosystems into the boundaries of the city to deliver UES. This has included ecological restoration which strives to recover native biota and all previous ecosystem functions. It has also included ecological rehabilitation and biological systems engineering which strive to reinstate, introduce and/or emulate ecosystem functions to a level where they deliver beneficial social, health, environmental, recreational, educational, and/or economic UES (Bush and Doyon 2019, Lehmann 2021). Nature-based projects have been used to rehabilitate and/or (re)introduce ecosystems, wetlands, forests and/or green infrastructure along riparian corridors and coasts, within large and small parks and open spaces, in spaces between buildings, and along street, rail and infrastructure corridors (Crowe and Rotherham 2019, van Leeuwen et al. 2019).

Biological systems engineering and stormwater management

The confined artificial spaces of cities are often difficult places for flora and fauna to survive and flourish, therefore, restoring natural ecosystems in urban areas is rarely possible without human assistance (Bush and Doyon 2019). As a result, bioengineering and ecosystem emulation have been used rather than restoration to develop socioecological systems that deliver sought after UES. Ecosystems emulation and bioengineering which combines flora and fauna with synthetic structures and elements, has been an integral part of greening cities, ecosystem rehabilitation, and nature-based projects (Garcia 2017, Pedersen Zari 2017). In many interventions, bioengineering adopts many of the principles of complex ecosystems to create new systems that combine adroitly chosen flora and fauna with synthetic materials and elements to deliver specific UES such as erosion control, flood mitigation, wildlife protection, and visual and acoustic protection (Garcia 2017).

Water sensitive urban design and stormwater management

An area where ecosystem emulation bioengineering has occurred has been the adoption of WSUD in stormwater management. WSUD has been developed and used to overcome some of the problems associated with traditional stormwater management. Traditionally, as cities have grown and developed, the natural hydrology of the land, its watercourses, and natural ecosystems were removed or modified to maximise the amount of land available for urban development (Crowe and Rotherham 2019, van Leeuwen et al. 2019). In the process, the soft surfaces of land and plants were progressively replaced by the impermeable surfaces of rooves, roads, carparks, and lawns, reducing the ability of the land and its biotic and abiotic elements to absorb rainfall. Without soft surfaces and plants to absorb and slow rainwater’s passage to rivers, seas, and aquifers, it causes torrents, inundations, erosion, and flooding (Crowe and Rotherham 2019, van Leeuwen et al. 2019). The mitigation of these problems was traditionally addressed with hierarchical engineered systems of kerbs, drains, subterranean pipes, and canals designed to transport stormwater out of the city as efficiently, safely, and quickly as possible (Crowe and Rotherham 2019, van Leeuwen et al. 2019).

A negative consequence of hierarchical engineered systems is they turn a valuable resource (water) into a toxic hazard. As they gather water from hard surfaces, they mix it with solid rubbish, road dust (compounded particulates from exhausts, tyres, and mechanical wear from motor vehicles), excrement from animals, chemical waste from industry, and fertilizers, herbicides and insecticides from lawns and gardens. The resultant toxic flow degrades the watercourses and waterbodies of the city, and then the land, seas, rivers, lakes, and natural environments beyond the city where the polluted water ends up (Frank and Engelke 2005, Ewing and Handy 2009). In addition, the evolving nature of the city, such as increasing hard surfaces and aging infrastructure, as well as changing weather patterns and rising sea levels due to climate change place mounting pressure on the present and future ability of many engineered stormwater systems to reliably carry out their designated role of preventing disruption and damage from flooding (Revi et al. 2014).

Water sensitive urban design and complexity

WSUD systems combine human-made elements and engineering with the hydrology of the land and ecosystems to emulate natural hydrological systems. To do so, they integrate human-made tanks, pumps, and pipes with biodiverse street level raingardens, wetlands, ponds, swales, rehabilitated creeks, rivers, and aquifers (Figs 2 and 3) to create complex systems that slow, filter, capture, store, and reuse stormwater (Crowe and Rotherham 2019, van Leeuwen et al. 2019). Plants, human-made apparatuses, synthetic materials, and abiotic elements such as rocks, stones, and selected soils are assembled in these systems to carry out ecosystem emulating functionalities that alone or in relationships with one another create diverse UES as emergent phenomena (Crowe and Rotherham 2019, van Leeuwen et al. 2019).

Corresponding with principles from natural complex systems, WSUD creates system mass via the use of multiple diverse and integrated small and slow solutions at multiple scales. It uses human created complex ecosystems at multiple levels to produce a complex human designed stormwater system, with resilience created by diversity, connection, linkage, and redundancy (Johnson 2002, Page 2010). In these systems, if one part of the system fails, the rest of the system can maintain functionality and can often adapt and compensate, therefore providing resilience via redundancy.

Adaptive management of socioecological systems

While human rehabilitated and bioengineered urban socioecological systems reflect principles of complex ecosystems, they do not usually create permanent self-organisation. In addition, while they are often complex in regard to connected diversity, symbiotic relationships, and emergent phenomena, they primarily exist to deliver specific and valued UES, therefore they are not open to adaptation or evolution in ways that diminish or lose these. Furthermore, bioengineered socioecological systems are recent additions and still novel in most cities, therefore, ongoing experimentation and change can be used to improve or expand UES delivery. Finally, urban environments are subject to constant exogenous phenomena such as pollution, human activities, exotic weeds and pests and climate change which makes them difficult places for flora and fauna to flourish and for ecosystems to maintain diversity and symbiotic relationships (Breuste et al. 2013, Gómez-Baggethun and Barton 2013, Bush and Doyon 2019). Therefore, they need ongoing human interventions (adaptive management) at multiple scales (Westgate et al. 2013, Krebs and Bach 2018, Sharifi and Yamagata 2018).

Adaptive management alternates between observation and interaction to constantly build knowledge about the system’s components and dynamics, then use this information to incrementally improve them (Westgate et al. 2013, Krebs and Bach 2018, Sharifi and Yamagata 2018). Adaptive management emulates the bottom-up endogenous dynamic found in natural complex adaptive systems. In natural systems, this dynamic occurs via mutation and adaptation over generations (Page 2010). In socioecological systems, it can be gained and implemented at any time via scientific discovery, trial, and error experimentation, and then spread widely and replicated as knowledge (Westgate et al. 2013, Crowe and Rotherham 2019, van Leeuwen et al. 2019).

Managing sociotechnical and socioecological transitions

WSUD systems have been used in new residential subdivisions and the regeneration of urban watercourses with beneficial UES; however, high costs, complexity, and need for multisector coordination and collaboration means they are yet to be achieved at a holistic metropolitan wide scale (Fletcher et al. 2015, Radcliffe 2019, Rashetnia et al. 2022). Stormwater management systems are complex socioecological and sociotechnical systems made up of land, infrastructure, and machinery, managed, synchronised, and regulated by multiple stakeholder organisations, investors, bureaucracies, suppliers, ancillary support providers and customers (Westgate et al. 2013). In addition, actor organisations within greater systems are controlled by workforces with established skill sets, experience, and cultures (Geels 2005, Köhler et al. 2019). There are also imbalances of power among stakeholders within these systems Therefore, a sociotechnical or socioecological transition from engineered solutions to nature-based solutions requires changes to technology, policy, markets, consumer practices, infrastructure, cultural meaning, skills, and scientific knowledge (Geels 2011, Garud and Gehman 2012, Ramos-Mejía et al. 2018). In addition, transitions can result in loss, disruption, and unintended consequences in other sectors, particularly where there is competition for limited resources such as land, time, or capital. Furthermore, individual organisations have their own strategies, work schedules, key performance indicators and budgets that may not coincide or even conflict with those of other organisations.

Due to these complexities sociotechnical systems often exhibit path dependency and when expedited or promoted by one area of government can face resistance from other sectors that slows or derails the transition process (Sharifi et al. 2017). As such transitions require, coherent long-term but adaptable responses in support of the new and a willingness to see incumbents lose. They require multisector coordination, commitment, and management, effective joined-up governance structures, as well as bundles of enforced policies (investments, taxes, incentives, regulations) (Innes and Booher 1999, Booher and Innes 2002, Geels 2011, Ramos-Mejía et al. 2018, McGreevy et al. 2020). They also require enduring adaptive management by multiple organisations to maintain optimal UES delivery long-term (Frantzeskaki 2019).

Complexity principles have been used to build better governance or collaborative planning processes and provide frameworks and principles for planners in their roles as negotiators, fixers, and brokers (Booher and Innes 2002, Chettiparamb 2013). In public administration scholarship, these processes have been referred to as horizontal governance or government, boundary spanning, or joined-up governance (Carey and Harris 2016). The argument is that in complex societies, where power information and knowledge are dispersed and the decisions of one actor can have intended and unintended effects upon other actors, there is a powerful need for collaboration in decision making processes. The goal of joined-up governance is to replace traditional compartmentalised hierarchical bureaucratic relationships competing for power and funding with enforced intersectoral collaboration that tackles issues spanning multiple sectors holistically (Ferlie et al. 2011, Albrechts 2016).

Effective joined-up governance requires latticed combinations of hierarchical and heterarchical connections within decision making systems that link a diversity of minds and knowledge. These systems emulate the reverberating information loops and subsequent collective intelligence dynamics found in natural complex adaptive systems. In joined-up governance systems dialogue and consensus are used to produce enforceable agreements (Innes and Booher 1999; Ferlie et al. 2011, Sørensen 2013). However, effective heterarchical integration can be problematic in public policy unless a single level of government has the jurisdictional power and broad responsibility across city regions to enforce the long-term relationship building, dialogue, and interaction required for collaboration and consensus-building (Newman 2008, Mossberger 2009, Reff Pedersen et al. 2010). Without the overarching organisation of an empowered government, important agents involved in urban development may refuse to participate fully and/or resist directions not perceived to be in their immediate interest (Innes and Booher 1999, Newman 2008, Reff Pedersen et al. 2010). Furthermore, Carey and Harris (2016) argue that like other complex social systems, joined-up governance systems can be disrupted and degraded, therefore, they require enduring adaptive management.

Singapore as a case study

Singapore is located in a biodiversity rich tropical molesian region (Chan 2019). While most of this biodiversity was lost in the colonial period, since independence there has been a long-term government led commitment to develop Singapore as a biophilic ‘lush tropical garden city’ (Tan and bin Abdul Hamid 2014, Friess 2017b, Han 2017, Chan 2019). In addition, the Singapore government views being internationally acknowledged as a ‘smart’ technologically and scientifically savvy global city as a key component of its national identity and status (CLC 2016). To this end, the nation has been highly successful and is routinely nominated as one of the smartest of smart cities (Chang and Das 2020).

The Singapore Green Plan is an overriding intersectoral ‘Big P’ plan for guiding and coordinating Singapore’s environmental policies and climate change adaption and mitigation policies. It contains policy commitments to expand and restore ecosystems and find and use scientific and nature-based solutions to overcome urban problems (MSE 2020a).

Singapore is hot and humid and subject to high rainfall often in significant deluges. It is predicted climate change will raise Singapore’s mean daily temperature by 1.4 to 4.6°C by the end of the century and rainfall will increase by ∼ 7% (MEWR 2016a, MSE 2020b). Rainfall is also forecast to occur in higher intensity downpours in the wet season, while there will be extended periods of low or no rainfall in the dry (MEWR 2016a). Furthermore, Singapore is a small, low-lying island. Therefore, it is vulnerable to coastal inundation from the increased storm activity and predicted sea level rises of between 0.25 and 0.75 m (MEWR 2016a). If left unchecked, these issues will lead to problems such as increased flooding, coastal land loss, water scarcity, loss of biodiversity, increased energy use, heat stress, and new diseases (MEWR 2016a).

Method

This article reports the findings of an analysis of the ‘Big P’ and ‘small p’ policies of Singapore government agencies with authority over sociotechnical and socioecological systems. The research aimed to investigate to what extent the ‘Big P’ policies contained in the Green Plan related to stormwater management and nature-based solutions are reflected in the ‘Big P’ and ‘small p’ policies of government departments and agencies with jurisdiction over land and resources. Furthermore, it investigated whether nature-based solution policies mentioned are cognisant of and/or incorporate complexity principles to produce emergent UES. Finally, it assessed whether the collected documents demonstrate a coordinated, holistic joined-up governance approach to policy formulation and implementation.

The documents were collected from the following nine departments and agencies ( Appendix 1):

• The Public Utilities Board (PUB) is responsible for stormwater infrastructure and management, reservoirs and their immediate catchments, coasts, as well as water distribution.

• The Housing Development Board (HDB) is responsible for spaces (soft & hard) between and around public housing buildings where 80% of Singaporeans live.

• The Ministry for National Development (MND) department with overarching authority over land use.

• The Urban Renewal Authority (URA) is responsible for the planning and management of major urban developments and redevelopments and manages vacant reserve urban land.

• The Ministry of Sustainability and the Environment (MSE), previously the Ministry of the Environment and Water Resources (MEWR).

• The Land Transport Authority (LTA) a subsidiary department of the MOT responsible for road infrastructure including the soft surfaces along road and street edges.

• The National Parks Board (Nparks) who are responsible for parks and gardens.

• The Building and Construction Authority (BCA) controls development approvals.

• The Ministry of Defence (MINDEF) is responsible for 19% of land designated for defence.

Document collection involved a researcher searching the websites of the aforementioned departments and agencies to collect current ‘Big P’ and ‘small p’ policy documents (including websites). The analysis included 14 ‘Big P’ and 15 ‘small p’ policy documents ( Appendix 1). We also analysed five documents from the Centre for Liveable Cities (CLC). The CLC is a division of the Ministry for National Development (MND) and has a stated mission to distil, create and share knowledge on liveable and sustainable cities with government departments and the public (CLC 2021b). As such, it is influential rather than directive over policy. CLC documents also provide extensive detail on the planning and development of urban projects. Documents were regarded as current if they were available on the department website at the time of the search (July 2022 to December 2022). The documents cover a broad range of topics, including urban design, land use planning, passenger transport, coastal planning, infrastructure, housing, parks and open spaces, climate change adaption and mitigation, economic development, and water supply.

The research used structured content analysis of documents in order to find and categorise explicit content. Researchers examined the documents to ascertain what is present and not present in the data, and to what effect. Researchers sought to identify and understand commonalities and differences across documents and agencies and whether there was explicit evidence of collaboration and holistic approaches across jurisdictions. The objectives of the analysis were first to see whether the ‘Big P’ policies of the Singapore Green Plan were consistently and holistically reflected in the ‘Big P’ and ‘small p’ policies of the government agencies and departments listed above. Therefore, it shows evidence of joined-up governance or collaborative planning approaches. The second objective was to observe the extent to which ‘Big P’ and ‘small p’ policies include the delivery of UES through complex adaptive ecosystem, restoration, rehabilitation, or emulation.

A second tier of the research was an observational analysis of recent developments across Singapore to determine the extent to which complexity ideas such as WSUD are being utilised and leading to tangible outcomes. The observational analysis was undertaken using a checklist of elements required in holistic water sensitive designs from the street to watercourses. The checklist was based on principles from scholarship and local guidelines (PUB 2018, Crowe and Rotherham 2019, Sharma et al. 2019, van Leeuwen et al. 2019). Using Google Earth, Street View and site visits researchers looked for WSUD infrastructure such as swales, rain gardens, detention ponds, and nature-based filtration as well as evidence of holistic adoption of WSUD infrastructure along streets, watercourses, in public open spaces and spaces between buildings in recent development and redevelopment projects. These included rewilding efforts at multiple scales, a large HBD development at Punggol, a URA planned housing project at Lentor Hills, and the Nparks/PUB Kallang River rehabilitation.

Findings

Policy document analysis

A theme which is present in national policy and evident across jurisdictions is that science, technology, and engineering are keys to solving current urban problems and building resilience to future problems from climate change (RIE 2020, CLC 2021a). In addition, the ‘Big P’ policies of most government departments include a commitment to improving the liveability of the city as well as mitigating climate change via nature-based solutions such as ecosystem restoration, rehabilitation, and bioengineering (URA 2012, MND 2013, MEWR 2015, PUB 2018, Nparks 2019, MSE 2020a, 2022a, RIE 2020, HDB 2023). There is also recognition in the plans of most agencies that their actions affect other sectors. However, three major departments, the Land Transport Authority (LTA 2019), the Building & Construction Authority (BCA 2022), and the Ministry of Defence (MINDEF 2023) fail to mention in their ‘Big P’ policies their potential contribution to the natural environment, the affects nature-based solutions might have in their jurisdictions, or the potential for intersectoral collaboration to deliver holistic UES.

Policy documents from multiple departments have several key targets related to water and biodiversity that replicate those in the Green Plan (Nparks 2019, MSE 2022a). The water target is to reduce average per capita consumption to 130 litres per day. The biodiversity target is to add one thousand hectares of new greenspace by 2035, this includes planting one million trees, developing 130 hectares of new parks, and increasing the naturalness and lushness of 170 hectares of existing parks by 2030 (MSE 2022a). Finally, there are complementary but untargeted objectives to mitigate the city’s vulnerability to the effects of climate change, this includes increasing the city’s resilience to flooding, coastal inundation, and rising temperatures.

There is broad recognition in intersectoral documents such as the Green Plan (MSE 2022a), and ‘Big P’ policy documents from HDB (2023), MND (2013), Nparks (2019), MEWR (2016a), MSE (2020a), PUB (2014) and URA (2023) that Singapore has the potential to benefit from urban ecological services derived from a transition to nature-based solutions in its socioecological systems. A notable exception is the LTA which does not mention nature-based solutions in their main ‘Big P’ policy document (LTA 2019) or the strong relationship between the quality of green infrastructure and active transport. The only climate related policy mentioned is to construct engineered canopies over sidewalks leading to MRT stations.

‘Big P’ policy documents from HDB (2013), MND (2013), Nparks (2019), MEWR (2016a), MSE (2020a), PUB (2014) and URA (2023) point to direct UES benefits from nature based rewilding and WSUD such as cleaner waterways and reservoir water, and reduced flooding, as well as emergent UES such as lower ambient temperatures, less air and noise pollution, and more interesting and attractive open spaces for recreation, socialising, and communion with nature. Some documents also mention further emergent UES such as improved mental and physical health and wellbeing, which in turn provide further benefits such as fewer medical visits, and improved concentration and attendance at school and work.

Rewilding

The rewilding efforts have strong links to the unique contribution the metropolitan area of Singapore makes to the water supply. 30%–40% of Singapore’s water supply comes from rainwater runoff from a catchment that covers 66% of the city (CLC 2020). This runoff is caught and then stored in 17 reservoirs within the urban fabric. The historical necessity of using the city as a catchment and storing water within the bounds of the city has seen Singapore designate land for catchments immediately around major reservoirs in the centre and northwest of the island (CLC 2020). These areas have historically contained a variety of low intensity uses, however, over the last few decades they have been the focus of rewilding (Chan 2019). As a result, the low intensity uses that once proliferated in these areas have been removed and natural ecosystems restored or rehabilitated (Fig. 4) via the reintroduction of indigenous flora and fauna as well as some complementary engineering (MEWR 2015).

Apart from water storage areas, the rewilding effort has also included ecosystem reintroduction, restoration, and rehabilitation in large national parks, and along stretches of some large water courses and infrastructure corridors (Nparks 2019, URA 2023). It has also included selective plantings in smaller parks to create ‘lush’ native gardens with some but not complete biodiversity (MND 2013, Nparks 2019, MSE 2020b, HDB 2023). In addition, there have been policies and projects using biophilic urban design to landscape smaller spaces between buildings and along watercourses, drains, canals, infrastructure corridors and roads (Nparks 2019, HDB 2023). Walking/cycling paths and recreational facilities have also been installed as part of these efforts. Rewilding along infrastructure corridors, and in small parks and spaces between buildings often lacks the species diversity, synergies and self-organisation required of self-sustaining complex adaptive ecosystems. However, they still provide some UES, such as shade, stormwater slowing and absorption, microclimate regulation, beauty, habitat for fauna, and recreation opportunities (Fig. 5).

Water sensitive urban design and stormwater management

A major intersectoral project that has advanced Singapore’s biophilic development is the Active, Beautiful, and Clean (ABC) water initiative which began in 2006 (PUB 2018). The ABC Waters initiative advocates a whole of catchment response to stormwater management incorporating WSUD principles and projects. As its name suggests, the program sees multisector benefits from the introduction of WSUD, active as opportunities for recreation, beautiful to enhance the city’s attractiveness to residents, prospective migrants, visitors, and investors; clean to improve the quality of water runoff entering watercourses and reservoirs (PUB 2018).

PUB and other agencies lead the ABC initiative with jurisdiction over large areas of the city’s water; the LTA, URA, Nparks, and HDB are signatories. However, the nation’s largest landholder, the Ministry of Defence, is not a signatory. Design guidelines (PUB 2019) provide comprehensive WSUD information for planning, designing, and constructing swales, rain gardens, green rooves, detention ponds, bioretention basins, water course rehabilitation and more. The ABC Waters initiative has also developed local WSUD expertise that can be called upon to assist in the design and deployment of WSUD in projects (PUB 2018). References to the ABC initiative appear in the ‘Big P’ policies of major government departments such as HDB (2023), MND (2013), Nparks (2019), MSE (2020b), PUB (2014) and URA (2023), as well as ‘small p’ policies such as the HDB Landscape Guide (HDB 2013), The Green & Liveable City Singapore Urban Design Guidebook (URA 2023), The BCA Green Mark for Infrastructure (BCA 2023), and the Innovate Collaborate Sustain Guide (MSE 2022a). However, incorporating ABC Waters guidelines into new projects is voluntary and is not always referred to in ‘small p’ guidelines and standards. For example, within the Walking and Cycling Design Standards (URA and LTA 2018), the guidelines for verges and stormwater infrastructure do not mention WSUD. Instead, they advocate engineered designs of kerbs, drains, pipes and verges with street trees and understories of cowgrass lawns.

The major strategy for the management of stormwater infrastructure and flood prevention is contained in the Sustainable Climate Action Plan (MSE 2020b). It advocates a Source, Pathway and Receptor approach. The objective of this approach is to increase surge capacity within the 8000 km of existing engineered canals and drains, and street level infrastructure that captures stormwater from the urban catchment and transports it to the city’s reservoirs (PUB 2014, MEWR 2016a, PUB 2018, MSE 2020b). This approach continues to rely on engineering with some room for nature-based solutions. The Pathway component relies primarily on engineered solutions such as enlarging, widening, and deepening existing infrastructure and constructing large detention tanks. The Receptor component consists of engineered responses such as building barriers and raising infrastructure above flood levels (PUB 2014, MEWR 2016a). The Source component is where nature-based solutions such as WSUD are most supported. All developments over 0.2 hectares are required to have interventions that slow runoff entering the stormwater system. These can be engineered detention tanks or nature-based solutions such as green rooves, detention ponds, swales, and rain gardens (PUB 2014).

Implementing rewilding and Water Sensitive Urban Design in projects

There were forty-three large and small ABC Waters projects completed up to 2020 including in schools, industrial estates, a hospital, and along water courses (CLC 2020). A further 100 ABC Waters projects have been identified for action out to 2030 (PUB 2018, CLC 2020). In addition, Nparks is incrementally retrofitting multi-layered plantings of trees shrubs, ferns, and grasses into road verges and median strips (Fig. 6). However, there is not enough species diversity to create a complex ecosystem and it is rarely integrated with PUB stormwater infrastructure to create WSUD infrastructure such as raingardens and swales. As a result, while the rewilded verges look attractive and absorb some stormwater runoff from sidewalks, much of the polluted runoff from road surfaces still flows straight into drains and subterranean pipes without much in the way of filtration, cleansing or slowing (Fig. 7). Therefore, while the active and beautiful components of ABC Waters are being achieved, the clean component (WSUD) is usually absent or not holistically installed.

There are also issues of ongoing maintenance preserving UES. Over the medium to long-term the ongoing adaptive management and maintenance of common property in HDB estates is the responsibility of underfunded Town Councils, not the departments who developed them. There are examples of complex gardens including WSUD infrastructure such as bioswales being removed by Town Councils and condominium managers and replaced by a more easily managed landscape of concrete and lawn.

Retrofitting nature-based solutions: the Kallang river project

There are two large projects that are highlighted in multiple ‘Big P’ policies and CLC documents as good practices in integrating ABC Water principles. They are Kallang River at Bishan/AMK Park and the LRA/HBD master planned new town of Punggol. Punggol’s approach to stormwater management is referred to as a model for future HDB and URA developments such as Jurong Lakes and Tengah (CLC 2021b).

The Kallang River project is an ecosystem emulating rehabilitation of a watercourse along the river’s upper reaches. In this project, a 2.7 km concrete drain was removed, and the watercourse was rehabilitated and integrated with public land along its banks to create a naturally flowing section of the river incorporating a public park (CLC 2019) (Fig. 8). This rehabilitated stretch has incorporated multiple WSUD interventions including bed layering, bioengineered riverbeds, swales and the reintroduction of native flora and fauna (Lim and Lu 2016). The UES projected to be gained from the project when under design included flood mitigation, drought amelioration, recreation, education, and increases in biodiversity (Chan 2019, CLC 2019). Today the park is a large detention pond to hold and slowly release stormwater from Singapore’s frequent tropical deluges and includes several recreation facilities such as walking and cycling trails, picnic areas, playgrounds, and sports facilities.

However, WSUD infrastructure or principles are not present in the park’s immediate lateral catchments, such as along feeder drains that are the responsibility of PUB, along adjacent streets and gutters that are the responsibility of the LTA, or within spaces between nearby buildings that are the responsibility of Town Councils and HDB. In addition, the rehabilitation abruptly ends when the Kallang River begins flowing along the side of LTA rail yards. From this point on the ‘river’ continues as a concrete canal (Fig. 9). There have been some nature-based interventions along the remainder of the river, such as revegetation along the canal itself and some drains emptying into it, however, these are ad hoc, not catchment wide or holistically integrated. In addition, a major renewal of the Ang Noi Kio town centre in 2024 in the Kallang catchment has not included WSUD (Figs 10 and 11).

Nature based solutions in broadacre developments: the Punggol new town and the Lentor Hills subdivision

Nature based projects have been more holistically incorporated into the broadacre development of Punggol. Punggol was initiated as Singapore’s 23rd new town in 1996. However, after a slow start, it was relaunched and redesigned as a biophilic waterway town in 2007. The Biophilic town framework was a joint project between PUB, HDB, Nparks, MND and URA. This included the incorporation of ABC (WSUD) Waters principles (CLC 2021b, MND 2013). On maturity, Punggol is expected to have a population of around 250,000 residents. In 2022, its population was approaching 200,000. The Punggol area had previously been a dumping ground for hard waste and ‘nightsoil’ and the location for noxious industries such as intensive piggeries and hatcheries. Consequently, the land in the area was contaminated and watercourses were polluted by runoff leaching from these facilities (CLC 2021b).

The project largely conforms to the Source, Pathway and Receptor response of combining engineered and biological mechanisms. At the receptor level, this included damming and deepening the Punggol and Serangoon Rivers at their mouths and connecting them with a 4.2 km ‘waterway’ canal (MSE 2020b, CLC 2021). Together these form a connected reservoir which is also a focus for urban development and resident’s recreation (Fig. 12).

Another engineered solution was the construction of an 18-metre deep, 6-km-long subterranean wall to keep contaminated water from leaching from the disused dump sites into the reservoirs. In addition, phytoremediation techniques have been used to help decontaminate land in these areas (CLC 2021b). Nature based WSUD solutions have also been used along the edges of the waterways and many pathways leading to them (Fig. 13). This has included wetlands of reeds, shrubs, grasses, and trees as well as the planting of mangrove forests along reservoir and watercourse edges (CLC 2021b). The reservoirs also contain human-made platforms that form the base for floating reed beds (Fig. 14). In the surrounding residential areas, the spaces between buildings are formed into parks that often incorporate features such as bioswales, bioretention basins, vegetated swales, cleansing biotopes, constructed wetlands and rain gardens. Many buildings also incorporate rooftop gardens (CLC 2021b). They act in a coordinated manner to slow and filter water from hard surfaces, lawns, and buildings. They also provide places of activity and natural beauty.

However, the road system and its drainage follow a traditional engineered form. There are wide bitumen roads flanked by pedestrian areas of concrete paths with stormwater from roads and paths draining from kerbs to drains and then into subterranean pipes (Fig. 15). Most verges are planted with immature trees and the ubiquitous cowgrass. There are some streets with more dense and complex understories of shrubs and grasses. However, they are rarely integrated with stormwater infrastructure.

A more recent greenfield development at Lentor Hills is following a similar pattern. The 48-hectare site was mostly cleared of the mature secondary forest that had covered it prior to development. The master plan incorporates a hilltop park in the centre of the subdivision retaining remnant vegetation (Fig. 16) and a linear park incorporating a degraded watercourse along the western flank of the development, which will include rewilding. In 2022, only the roads and infrastructure had been constructed. The stormwater infrastructure at this stage consists of traditional kerbs and subterranean pipes along the road system (Fig. 17). There are currently no WSUD features such as swales or rain gardens along these street corridors.

Other nature based bioengineered projects

There are also non-stormwater related projects that include engineered, bioengineering and nature-based solutions to adapt to the present and future effects of climate change. There are initiatives across the built-up areas of the city to install and retrofit green walls and rooves to mitigate heat island effects in built up areas (HDB 2013, URA 2023). Engineered solutions are being installed to mitigate rising sea levels and intensity from storm surges such as constructing sea walls, tidal gates, groynes, and pumping stations (MEWR 2016a). The responses to rising sea levels and storm surges have also included the reintroduction and/or rehabilitation of forty hectares of mangrove forests using biological systems engineering and natural ecosystem emulation (MSE 2020b). These projects combine synthetic elements such as biodegradable bags, introduced rocks, and PVC pipes and pots with the poly planting of adroitly selected plant species to reconstruct permanent forest systems in often difficult locations (MEWR 2016a, Nparks 2019, MSE 2022a). In the documents supporting these projects, the multiple UES benefits of mangrove forests are recognised. These include carbon sequestration and storage, wave attenuation, pollution and sediment trapping, habitat for diverse fauna, and places of spiritual and heritage importance (MSE 2022a). However, the forty hectares rehabilitated so far is a small fraction of the approximately seven thousand hectares that have been lost to agriculture, urban development, land reclamation, and reservoirs since 1820 (Friess 2017a). Another bioengineered pilot project underway is constructing offshore reefs as a means of protecting Singapore from storm surges. This initiative uses human-made modular reef structures to replicate a reef slope as a substrate for native corals and to provide niches for other marine organisms. This project is still in its formative stages and just 1000 m² of reef has been established so far (MSE 2020b, 2022a).

Discussion

The smart city ethos of Singapore, its global significance as a research and development hub, its historical commitment to biophilic urban planning philosophies, and the use of the city environs as a water catchment corresponds well with the science, technology, bioengineering, and ecosystem emulation required for ecosystem restoration and rehabilitation. Furthermore, because a single level of government both owns and regulates 90% of the land in Singapore, there is scope for coordinated holistic approaches. Therefore, rewilding and bioengineering are initiatives that expand upon well-established paths rather than necessarily transitions requiring shifts in path dependency, vested interest, experience, expertise, skills, and culture.

The ‘Big ‘P’ strategic documents and plans of multiple departments recognise the benefits of rewilding and bioengineering and nature-based solutions. There have been tangible results in rewilding efforts, particularly around reservoirs where natural ecosystems have been scientifically restored or rehabilitated. These projects have occurred under the jurisdiction of two collaborating government departments PUB and Nparks. Government funding and scientific resources have also been pumped into research and pilot projects that use bioengineering as a means of better managing urban infrastructure to produce beneficial UES as well as build resilience in the face of risks associated with climate change such as rising sea levels, heavier rain, storm surges, prolonged droughts, and hotter weather. There is also wide recognition in multiple documents from multiple departments that environmental initiatives can provide emergent intersectoral social and economic UES.

The greatest and longest bioengineering initiative so far has been the adoption of WSUD via the ABC Waters program. ABC Waters’ documents demonstrate intersectoral awareness of the benefits of WSUD as well as detailed design guidelines. The initiative has also developed expertise to implement WSUD holistically across jurisdictions. Nevertheless, many of the obstacles that can slow or derail a sociotechnical or socioecological transition described in transition management theory are present in the management and planning of spaces and infrastructure that form Singapore’s stormwater system. Stormwater management inevitably competes with existing uses for land and resources where it can encounter resistance. Engineered stormwater drains and canals have traditionally used depth, straightening and flow rates to prevent flooding, therefore, the introduction of WSUD into stormwater systems almost unavoidably requires widening at the expense of established uses (Sharifi et al. 2017). It also requires indefinite maintenance and adaptive management (Frantzeskaki 2019).

However, the volunteerism of the ABC Waters policies means an enforceable agreement reached by consensus required for a successful long-term transition is absent (Jørgensen et al. 2017, Rogge et al. 2017). Intersectoral commitments at the ‘Big P’ policy level and in the ABC Waters program, have not holistically trickled down into the ‘small p’ regulations, guidelines, and projects of departments overseeing land and infrastructure associated with WSUD. As a result, WSUD in Singapore is proceeding as ad hoc projects often chosen because they represent the paths of least resistance or least contestation not as parts of a holistically planned and coordinated city-wide transition. An absence of WSUD principles is observable in Nparks led ‘rewilding’ of road corridors. The current program of replacing cow grass lawns with trees and biodiverse understories is relatively inexpensive, requires no diversion of land from other uses and provides positive UES such as minor slowing and cleaning of stormwater as well as improving the attractiveness, safety, and coolness of the public realm. However, incorporating plantings with new WSUD stormwater infrastructure would have a significantly greater effect on slowing, filtering, and cleansing stormwater runoff from a wider catchment. It would also be more expensive, require the annexation of road space, and need ongoing adaptive management.

The Kallang River rehabilitation is an example of an incomplete WSUD project. The rehabilitation of the upper reaches of the river in AMK-Bishan Park has been widely feted and has improved the attraction, beauty, and cleanliness of the river (Lim and Lu 2016, Irvine et al. 2020). The project required the removal and rehabilitation of a concrete drain and the subsequent watercourse’s integration with an existing area of open space; therefore, no land needed to be diverted from other uses. Despite the successes of the initial project, the rehabilitation has not extended any further downstream or laterally into the river’s catchment. Rehabilitating the next 2-km downstream length (Fig. 11) would require the relocation and demolition of rail yards at a considerable financial cost and inconvenience to the LTA. Extending WSUD laterally into the catchment would also come at a considerable expense of capital and land. Integrating rewilding efforts with WSUD infrastructure would require the replacement of kerbs, drains, and pipes with raingardens and swales which would also often require the annexation of LTA managed road space and HDB owned and Town Council managed spaces between buildings. It would also require ongoing maintenance and adaptive management coordinated between multiple agencies.

Retrofitting WSUD is inherently difficult due to legacy form and path dependency, incorporating it into large greenfield developments such as Punggol and Lentor Hills is far less challenging. Therefore, it is surprising that the WSUD approach advocated in the ABC Waters guideline has not been holistically incorporated into either of these developments. ABC Waters guidelines have been included in some parts of Punggol and have proven successful at improving water quality (Yau et al. 2017). However, it is not holistic and there is a notable absence of WSUD infrastructure along road corridors. Lentor Hills similarly lacks WSUD infrastructure along streets and was planned and subdivided well after the launch of ABC Waters guidelines. Furthermore, the absence of WSUD features in the Ang Mo Kio town centre renewal demonstrates an absence of a whole of government approach to the transition.

Singapore’s transition to nature-based systems in general and WSUD in stormwater management, in particular, reflects the national ethos but is still slowed by competing claims, priorities, and path dependency. The absence of coordinated activity suggests the joined-up governance approaches derived from complexity theory are absent or at least undeveloped in policy implementation systems (Booher and Innes 2002, McGreevy et al. 2020).

Finally, rewilding efforts have not occurred without controversy. A result of the success of rewilding efforts so far has been the growth of fauna in the rewilded areas spilling into built up areas. On the positive side, the sighting of rare birds and animals in residential areas has produced excitement. On a more negative note, it has also led to distressing roadkill, otters devouring expensive ornamental pond fish, and macaques and wild boar attacking bins and occasionally people (Gerstein 2021, Iau and Sundar 2021, CNA 2022). However, there are no reports of rewilded areas becoming problematic habitats for disease vectors such as mosquitoes.

Conclusion

To a significant extent in ‘Big P’ policy and to a lesser extent in ‘small p’ standards and projects Singapore has purposely adopted nature-based principles into some socioecological and sociotechnical urban systems to deliver urban ecological services (UES). The program of nature-based solutions and using science and technology to enhance these as bioengineering has a historical legacy to build upon.

There are a number of restoration and rehabilitation projects as well as bioengineering projects that incorporate ecosystem principles and natural elements to produce diverse UES. The ABC Waters program has used ecosystem emulation and WSUD principles to develop comprehensive guidelines and expertise to holistically roll out WSUD across the city and, therefore, help facilitate a transition from an engineered system to a nature based complex system. However, the adoption of WSUD in large or small projects across the city remains voluntary. There is also a notable lack of coordination between departments in the rolling out of nature-based solutions. In particular, the LTA could at best be seen as a half-hearted partner in the ABC initiative. As a result, the adoption of WSUD and nature-based solutions more generally is currently progressing via pilot schemes and ad hoc projects in places of least resistance often providing the active and beautiful component of ABC Waters but not the clean.

Despite its ethos of biophilia and strengths in science technology and research and development, strong central government, and interdepartmental policies such as the Singapore Green Plan, many of the obstacles to holistic coordinated effective transitions raised in transition planning scholarship seem to be in place in Singapore. Nevertheless, the pilots and projects already undertaken and planned provide examples to inform more holistic transitions both in Singapore and for other cities in Southeast Asia and beyond with similar geographies and climates but that lack Singapore’s considerable research and development resources. The challenge remains to develop governance structures and funding arrangements that can expand these as part of a coordinated holistic island wide transition.

Author contributions

Michael Patrick McGreevy (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Investigation [equal], Methodology [equal], Writing—original draft [lead], Writing—review and editing [lead) and Eng Seng Chia (Aaaron) (Project administration [equal], Writing—review and editing [supporting])

Conflict of interest: None declared.

Funding

None declared.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Appendix 1: Government documents

References