1 Introduction

The human population has been growing exponentially, leading to a rapid expansion of cities. Currently, we have exceeded 8 billion people, with projections indicating that we will surpass 9.7 billion by 2050 (UN 2022a). Urban growth has been particularly pronounced in coastal areas of developing countries (Firth et al. 2016; Merkens et al. 2016; Reimann et al. 2023). Approximately one-third of the global population (2.75 billion) already lives within 100 km of the coast. This figure rises to almost half of humanity (3.5 billion) when considering people living within elevations up to 100 m above sea level (Reimann et al. 2023). However, population growth – and consequently urbanisation – results in one of the most irreversible environmental impacts, being one of the major causes of habitat loss and biodiversity decline (McKinney 2002, 2006, 2008; Grimm et al. 2008; Todd et al. 2019).

Coastal areas not only hold disproportionately high populations but also face increased vulnerability to risk hazards intensified by climate change (McGranahan et al. 2007; Nicholls et al. 2008; Oppenheimer et al. 2019). Worldwide, coastal cities have been increasingly facing flooding and erosion as a result of storm surges, wave impact, and sea-level rise. Such events threaten the persistence of entire communities, causing enormous socioeconomical and environmental losses (Lincke and Hinckel 2021; Paprotny et al. 2021). Human settlements in low-elevation coastal zones (LECZ) – areas situated within elevations up to 10 m above sea level (McGranahan et al. 2007) – are particularly under increased risk hazards. Currently, about 900 million people live in LECZ (Reimann et al. 2023), with the population expected to continue growing throughout this century (Neumann et al. 2015; Merkens et al. 2016; Reimann et al. 2023).

As a result of urban expansion, natural coastal habitats have been progressively replaced with or altered by the construction of rigid artificial structures (AS) at the land-sea interface. This process is referred to as coastal hardening (Dugan et al. 2012). These AS include infrastructure for coastal protection (e.g., seawalls, breakwaters, groynes) and for port and nautical activities (e.g., quays, wharves, jetties, pontoons) (Bulleri and Chapman 2010; Gittman et al. 2015), built for the purpose of accommodating human needs without considering environmental sustainability. In many regions, all around the world, extensive portions of shorelines – often more than 50% – have already been replaced with or modified by AS (e.g., Gittman et al. 2015; Lai et al. 2015; Firth et al. 2016, 2024; Aguillera 2018; Floerl et al. 2021). To meet various urban demands and address the consequences of climate change, the construction of AS along coastlines is expected to increase substantially in the coming years, despite the impacts posed to the natural environment (Bugnot et al. 2021; Floerl et al. 2021).

Concern regarding the environmental impacts of AS on coastal habitats emerged in the 2000s, becoming a topic of great interest for science and society in the face of rapid urban sprawl. Studies ever since have demonstrated that AS support lower diversity and abundance of species than natural habitats (Chapman 2003; Gittman et al. 2016), facilitate the establishment of exotic species (Dafforn et al. 2012; Airoldi et al. 2015) and consequently cause changes in the composition of assemblages (Chapman and Bulleri 2003; Pardal-Souza et al. 2017) and the homogenization of biota (Bulleri and Chapman 2010). Furthermore, AS have been shown to alter ecosystem functions (Mayer-Pinto et al. 2018, Martinez et al. 2022a, b), ecological connectivity (Bishop et al. 2017), and energy transfer (Malerba et al. 2019). Preserving biodiversity and ecological functions of the marine environment is vital to maintain the valuable goods and services they provide, upon which society depends, such as fisheries, water quality, and climate regulation (Barbier et al. 2011; Austen et al. 2015).

As in most countries, anthropogenic pressure on the Brazilian coast is also elevated. Based on the last demographic census carried out in 2022, around a quarter of the population – 48.2 million people – lives in coastal municipalitiesFootnote 1 which represents only 5% of the national territory. When considering inhabitants living within 150 km from the ocean, this includes more than half of Brazil’s population (~110 million people) (IBGE 2022). Out of 17 Brazilian states with access to the sea, the capital of 14 is situated on the coastal territory, while two are nearby (Fig. 1). As of 2018, economic activities linked to the ocean and coast were estimated to provide more than 20 million jobs and contribute to nearly 20% of Brazil GDP (Carvalho 2022). However, as a result of long-term unsustainable occupation and usage the Brazilian coastal zone has been facing increasing environmental deterioration and loss of natural capital (Martínez et al. 2007; Martinez et al. 2022a, b). In recent years, political instability and poor environmental governance have worsened such panorama (Barbosa et al. 2021; Menezes and Barbosa Jr. 2021).

Fig. 1
figure 1

Map of Brazil showing: A Brazil and its states (light grey) within South America context; B the coastal and inland territory of Brazil, depicting the capitals of the 17 states with access to the sea; and (C) the coastal territory of the state of São Paulo, showing its regions and municipalities

Full size image

Among 26 states and a federal district, the state of São Paulo (SP) is the most populous and developed in Brazil. Over 44 million people live in this state – a population similar to Spain –, nearly 22% of the national population (IBGE 2022). Economically, SP contributes to 30-35% of the whole annual country GDP, being the largest economy in the whole South Atlantic. Much of its economic importance is linked to the biggest national port and industrial complex, both located on the coastal zone of the state. About 2.2 million people live on the SP coast, a figure that increased by 15.7% between 2010 and 2022 – the last two demographic censuses – while the populations of the SP state and Brazil increased by 7.6 and 6.5%, respectively (Table 1). Despite not being particularly populous, the SP coast is within 100 km from the São Paulo Metropolitan Region, one of the biggest urban agglomerations of the world with over 20 million inhabitants (IBGE 2022). As a result, the SP coast suffers with mass tourism, receiving millions of people during holidays and summer and experiencing strong real estate speculation driven by second-home market. Moreover, the SP coast has been increasingly facing climate-induced events and erosion (Rodríguez et al. 2016; dos Santos and Serrao-Neumann 2018; Nunes et al. 2019), causing socioeconomic and environmental losses and posing complex governance. Therefore, the SP coast is a strategic territory where its resilience to climate change will depend on mitigating impacts and adapting cities in response to unfolding changes. Overcoming such challenges relies on policy reforms aiming at sustainability and grounded on solid scientific knowledge.

Table 1 Population, area, demographic density, and population growth between 2010 and 2022 of Brazil, coastal municipalities of Brazil, the state of São Paulo (SP), the SP coast, and the regions of SP coast. These data were extracted from IBGE 2022 population census. M = million; k = thousand
Full size table

Here, we identified and mapped the extent of (i) natural habitats, (ii) artificial structures, and (iii) occupations in low-elevation coastal zones near marine environments along the coastline and within estuaries of the state of São Paulo. This study provides the first large-scale detailed inventory of such metrics of coastal urbanisation in this state, in Brazil, and in the southwestern Atlantic. The baseline dataset generated in this research delivers valuable information for diverse stakeholders (e.g., policymakers, decision-makers, and engineers) to guide better coastal management towards achieving sustainability and climate resilience.

2 Material and methods

2.1 Study area

The territory of the SP coast is generally delimited by an abrupt change in elevation due to the Serra do Mar mountain range. The municipalities are squeezed into a narrow land strip between the sea and the mountain, mostly not wider than 20-30 km. The SP coast is formed by 16 municipalities divided into three regions – south, central, and north (Fig. 1C). Comprising 3 municipalities, the south region is the least populated, holding only 2.5% of the inhabitants of the SP coast. This area is characterised by extensive government-protected areas of Atlantic rainforest, and the economic activities are linked to tourism, fishing, and small-scale aquaculture and agriculture. The central region is the most populous and urbanised, with a population of ca. 1.8 million people (82% of inhabitants of the SP coast) distributed across 9 municipalities. This area has municipalities with over 300,000 inhabitants, and the main economic activities are linked to the Port of Santos – the biggest and busiest in Brazil – and the industrial complex of Cubatão – also the biggest in the country. This region has a long history of human occupation and usage, which resulted in intense urbanisation and environmental degradation (Moraes 2007; Lana et al. 2018). Lastly, the north region is formed by 4 municipalities and holds 16% of the population of the SP coast (Table 1). The main economic activities are linked to the Port of São Sebastião and a petrochemical terminal, in addition to tourism, fishing, and small-scale aquaculture. The south region is characterised by a coastal plain, with the occurrence of estuarine complexes bordered by mangrove forests, while the shoreline is dominated by dissipative sandy beaches (Lana et al. 2018; de Mahiques et al. 2016). In the north region, the intricate coastline is predominantly formed by rocky shores, with numerous and usually smaller reflective sandy beaches (de Mahiques et al. 2016; Pardal et al. 2023). The central region, in turn, shares the features of the other regions, being characterised by areas with estuarine habitats and dissipative beaches interspersed with rocky shores and reflective beaches.

2.2 The extent of coastal hardening in SP

We measured the linear extent, in metres, of natural habitats and artificial structures (AS) along the SP coastline and within estuaries through aerial images from Google Earth Pro software using the ‘path’ tool. AS were classified into infrastructure running along the coastline (ASc, e.g., seawalls, breakwaters) or extending out from the shoreline into adjacent waters (ASe, e.g., jetties, pontoons, groynes). ASc represent the amount of modified/armoured coastline and are a proxy for the extent of habitat loss. ASe, in turn, represent the amount of artificial hard substrates that were added to coastal habitats. Coastal hardening is represented by the combined ASc and ASe. AS were measured at ~50-100 m parallel to the ground. AS less than 5 m wide were measured by quantifying their extent drawing a line in their central area, while wider ones were contoured. For images of homogeneous landscapes, such as beaches, mangroves, and rocky shores, mapping was carried out up to 500 m from the ground. The limit of the coastline mapped inland was delimited by mangrove areas and river/estuarine margins with a width of less than 30 m. For gentle-slope waterfront areas, such as beaches, we visually identified the upper limit of high tides and used it to delimit the coastline contours. The coastal zone, in the marine environment, was demarcated within 12 nautical miles (i.e., 22.22 km) from the continent, following the legal delimitation of the United Nations Convention on Law of the Sea. Islands within this distance were also considered.

AS were further classified into specific types, adapted from Floerl et al. (2021): (i) coastal protection breakwalls (i.e., structures such as seawalls, breakwaters, and groynes, generally built with concrete, rocks or a mixture of them); (ii) wharves (i.e., large structures commonly composed of pillars, reinforced with steel plates and covered with concrete, used for port activities and for docking large vessels); (iii) jetties (i.e., smaller structures for docking small vessels and used for leisure, commonly made of wood or concrete supported by pillarsFootnote 2); (iv) pontoons (i.e., structures with floating bases connected to fixed structures on land, generally made of concrete-coated polystyrene or polyethylene blocks); (v) buildings (i.e., overwater houses and other facilities supported by wooden or concrete pillars); and (vi) fishing and aquaculture apparatus (i.e., fish weir, mussel ropes, and oyster racks) (Supplementary Material, Fig. A1-A6). AS extent data were analysed for the entire SP coast, regions, and municipalities.

2.3 The extent of occupations in low-elevation coastal zones near marine environments in SP

Using the ‘path’ tool from Google Earth Pro software, we measured the extent of occupations in low-elevation coastal zones near marine environments (LECZ100m) along the coastline and within estuaries of SP. This was done by quantifying the linear extent of the coastline, in metres, with inland human settlements and urban infrastructure (e.g., cities, villages, highways) situated perpendicularly within 100 m away from marine environments and up to 5 m above sea level. These measurements did not include the coastline extent already replaced by artificial structures (i.e., ASc), although these are also situated in LECZ100m. The extent of coastline with inland occupations in LECZ100m provides data on areas with human settlements and assets under risk hazards and likely to be hardened in the future. LECZ are defined as areas with an elevation up to 10 m above sea level. Here we adopted an elevation up to 5 m above sea level to provide a more conservative scenario in consideration of estimative errors in Google Earth elevation data (average error = 0.51 m, accuracy = 1.85 m; El-Ashmawy 2016). Likewise, the 100 m inland horizontal delimitation considers conservative projections of shoreline retreat (Vousdoukas et al. 2020; Lansu et al. 2024). Moreover, lowland coastal areas in SP are under eminent risk hazards of, and already facing, flooding and erosion (e.g., Marengo et al. 2017; dos Santos and Serrao-Neumann 2018; Nunes et al. 2019). Under the described conditions, areas nearby sandy beaches and estuarine/river margins were included (Supplementary Material, Fig. A7). Rocky shores were not considered in this analysis because they are steep and provide natural coastal protection. The extents of the coastline with inland occupations in LECZ100m from sandy beaches and estuarine/river margins were analysed for the entire SP coast, regions, and by municipalities.

We also quantified the total extent of the soft coastline as the extent of estuarine/river margins plus sandy beaches, both with natural habitats and occupations in LECZ100m. Based on that, we calculated the percentage of the relative susceptible coastline as ((ASc + LECZ100m) / soft coastline)) × 100. This metric indicates the proportion of the coastline with adjacent human settlements and urban assets in lowland areas under risk hazards. We further quantified the total extent of natural areas for potential future hardening, i.e., total hardenable coastline, which was calculated as the sum of the extent of the coastline with natural habitats and with occupations in LECZ100m from estuarine/river margins. Sandy beaches were not used in this calculation because such habitats are unlikely to be replaced by AS at the intertidal zone. The percentage of hardened coastline considering the total hardenable coastline, i.e., relative hardened coastline, was calculated as (100 × ASc) / (ASc + hardenable coastline).

3 Results

3.1 General patterns of coastal hardening in SP

The total extent of artificial structures (AS) along the SP coast spans 244 km. From this total, 125 and 119 km correspond to AS running along the coastline (ASc, e.g., seawalls, breakwaters) and extending from the shoreline into adjacent waters (ASe, e.g., jetties, pontoons, groynes), respectively. 63% of the total extent of AS (154 km) is located in the central region. The north and south regions account for 22% (54 km) and 15% (36 km) of the remaining AS extent, respectively. In the central region, ASc (61%, 94 km) are more abundant than ASe (39%, 60 km). ASe, on the other hand, are more abundant in both the north (68%, 37 km) and south regions (62%, 22 km) (Table 2, Fig. 2A). The municipalities of Guarujá (56 km), Santos (49 km), and São Vicente (26 km) in the central region contain the greatest AS extents. For these municipalities, ASc are more common than ASe, ranging from 57 to 68% of their total AS extent. São Sebastião (19 km) and Ilhabela (15 km), in the north region, and Cananéia (19 km), in the south, are other municipalities with the following largest AS extents. For them, ASe are more common than ASc, ranging from 58 to 75% of their total AS extent (Fig. 2B).

Table 2 Summary of the extent of different features and habitats mapped along the coastline and within estuaries of the state of São Paulo (SE Brazil) and its regions. ASc = artificial structures running along the coastline; ASe = artificial structures extending from the shoreline into adjacent waters; LECZ100m = low-elevation coastal zone within 100 m from marine environments
Full size table
Fig. 2
figure 2

Linear extent of artificial structures running along the coastline (ASc, dark colours) and extending from the shoreline into adjacent waters (ASe, light colours) in the state of São Paulo (SE Brazil). Data are shown for regions (A) and municipalities (B)

Full size image

3.2 AS types mapped in SP

Breakwalls are the most common AS type found on the SP coast, amounting to a total linear extent of 108 km, followed by jetties and wharves (~40 km each), and aquaculture and fishing apparatus (~24 km). Pontoons (18 km) and buildings (12 km) are the least common types (Fig. 3A). Breakwalls and wharves are concentrated in the central region, where they amount to an extent of 70 and 34 km, respectively. Moreover, there are 22 km of buildings and pontoons in the central region (~10 km each), while aquaculture and fishing apparatus are absent (Fig. 3B). In the north region, breakwalls (23 km) were also the most abundant AS, followed by aquaculture and fishing apparatus (11 km). Jetties (~8 km), wharves (7 km), and pontoons are also common (~5 km), but not buildings (~0.2 km) (Fig. 3C). In the south region, breakwalls (~14 km), aquaculture and fishing apparatus (13 km), and jetties (7 km) are the most common AS. Wharves are absent, while pontoons (1.5 km) and buildings (~0.5 km) are less abundant (Fig. 3D).

Fig. 3
figure 3

Linear extent of different types of artificial structures running along the coastline (ASc, dark colours) and extending from the shoreline into adjacent waters (ASe, light colours) in the state of São Paulo (SE Brazil). Data are shown for the entire coast (A) and region (B-D). Note that the numerical axes are in different scales and were broken for better visualisation

Full size image

3.3 Occupations in low-elevation coastal zones near marine environments in SP

Over 300 km of the SP coast has occupations and urban infrastructure in low-elevation coastal zones near marine environments (LECZ100m) (Table 2). From this total, 235 and 67 km are situated within 100 m from sandy beaches (Fig. 4) and estuarine/river margins, respectively (Fig. 5). Considering all habitats, the central region amounts to an extent of 149 km of the coastline with occupations in LECZ100m, followed by the north (108 km) and south regions (45 km). Relative to the extent of the soft coastline, the susceptible coastline is greater in the north region (77%), followed by the central (~42%) and the south (6%) (Table 2).

Fig. 4
figure 4

Linear extent of the coastline with natural areas (dark colours) and urban occupations in low-elevation coastal zone within 100 m (LECZ100m, light colours) from sandy beaches in the state of São Paulo (SE Brazil). Data are shown for regions (A) and municipalities (B)

Full size image
Fig. 5
figure 5

Linear extent of the coastline with natural areas (dark colours) and urban occupations in low-elevation coastal zone within 100 m (LECZ100m, light colours) from estuarine and river margins in the state of São Paulo (SE Brazil). Data are shown for coastal regions (A) and the sixteen municipalities (B). For better visualisation, the axes in A and B were broken and LECZ100m data are shown separated (C)

Full size image

3.3.1 Occupations in LECZ100m from sandy beaches

The mapped extent of sandy beaches on the SP coast amounts to ~423 km, of which 55% (~235 km) has occupations in LECZ100m. While the total extent of sandy beaches is similar among the three regions (i.e., ~140 km each), the sandy beaches from the central and north regions are already massively urbanised: 81% (117 km) and 74% (100 km) of their total extent have occupations in LECZ100m, respectively. In the south region, the pattern is contrary: 77% (125 km) of the total extent of sandy beaches have natural formations in LECZ100m (Table 2, Fig. 4A). From the 8 cities with sandy beaches in the central region, 6 of them have 85-100% of their extent with occupations in LECZ100m. Peruíbe (54%) and Bertioga (62%) were the municipalities with less occupation in LECZ100m (Fig. 4B). The sandy beaches from the municipalities in the north region also have high degree of occupation in LECZ100m, ranging from 63 to 89% (Fig. 4B). Finally, in the south region, the municipality of Ilha Comprida amounts to 21% (~14 km) of the extent of sandy beaches with occupations in LECZ100m, while Cananéia and Iguape have only 3% (~1 km) and 5% (~2 km) (Fig. 4B).

3.3.2 Occupations in LECZ100m from estuarine and river margins

We mapped 1,250 km of coastline near estuarine and river margins on the SP coast, which are situated in the south (63%, 787 km), central (35%, 437 km), and north regions (2%, 26 km). From the total mapped extent, 95% (1,183 km) still have natural formations (e.g., mangrove, rainforest, mudflats) in LECZ100m: 760 km (97%), 405 km (93%), and 19 km (72%) in the south, central, and north regions, respectively (Table 2, Fig. 5A and 5B). Occupations and urban infrastructure in LECZ100m occur in the remaining 67 km (5%) of the coastline, most of it in the central (48%, 32 km) and south (41%, 27 km) regions, with only 7 km in the north (Fig. 5C). Moreover, Iguape (~14 km), in the south, and Santos (~9 km), in the central region, are the municipalities with the largest extents of the coastline with occupations in LECZ100m from estuarine and river margins (Fig. 5C). The total extents of hardenable coastline in the south, central, and north regions are ~787, 437, and 26 km, respectively. The relative hardened coastline is greater in the north region (40%), followed by the central (~18%) and the south (~2%) (Table 2).

4 Discussion

The results of our study indicate that coastal urbanisation, in the form of coastal hardening and occupations in lowlands, is a major yet largely overlooked source of environmental impact and risk hazard in SP and Brazil. We have shown that the SP coast has a total linear extent of 244 km of hard artificial structures (AS), with 63% of them occurring in the most urbanised region. Breakwalls, jetties, and wharves represent 78% of these structures. AS running along the coastline (ASc, e.g., seawalls, breakwaters) accounts for 9% (125 km) of the ‘hardenable’ extent of the SP coast (Table 2). Additionally, our study revealed that 301 km of the SP coast has inland occupations in low-elevation coastal zones within 100 m from marine environments (LECZ100m). When considering ASc – which are also situated in LECZ100m –, the combined extent amounts to 427 km. This indicates that over 25% of the SP soft coastline has adjacent settlements and urban assets in areas susceptible to erosion and flooding (Table 2).

4.1 Coastal hardening in SP

The clustering of coastal infrastructure around populous and port cities is a global pattern (e.g., Dafforn et al. 2015; Firth et al. 2016, 2024; Floerl et al. 2021; Claassens et al. 2022). Accordingly, coastal hardening is advanced in the central region of the SP coast, which concentrated 63% (154 km) of the total extent of AS measured in this study. The extent of ASc alone (94 km) represents nearly 18% of its ‘hardenable’ coastline (Table 2). The central region has a long history of occupation associated with the construction and expansions of the Port of Santos and early industrialisation (Moraes 2007). Moreover, municipalities in this region have settlements historically situated on floodable areas and near eroding shorelines, which have propelled the introduction of coastal defence infrastructure (Souza et al. 2019). The majority of AS measured in the central region were breakwalls and wharves (68%, 104 km). This suggests that the proliferation of AS primarily arises from port activities and coastal armouring (Supplementary Material, Fig. A8 and A9). Considering projections for sea-level rise and increasing flooding risks (Marengo et al. 2017) alongside the Port of Santos' development strategy to double its land area (SPA 2020), it is expected that shoreline armouring and port operation will increasingly contribute to coastal hardening in the central region.

In the south and north regions of the SP coast, breakwalls were also the most common AS, although their extents were 3-5 times less than in the central region. The presence of these infrastructures beyond urban centres reflects past engineering interventions to stabilise and protect sedimentary coastlines and facilitate sea access (Bugnot et al. 2021). More recently, AS such as groynes, seawalls, and breakwaters have been built in response to increasing coastal erosion and flooding (Rodríguez et al. 2016; Barros et al. 2021) (Supplementary Material, Fig. A10). In the south and north regions, fishing and aquaculture infrastructure ranked as the second most widespread AS, and the most prevalent type extending from the shoreline into adjacent waters (ASe), adding 24 km of hard constructions into the sea. Such results are associated with the history of artisanal fishing and small-scale mussel/oyster farming in these regions, where water quality seems to be more appropriate for such activities. We identified three types of fishing and aquaculture apparatus: fish weirs (69%, 16.4 km), mussel ropes (18%, 4.2 km), and oyster racks (14%, 3.2 km) (Supplementary Material, Fig. A6). It is important to note that the reported measurements are probably conservative estimates, as they derive from satellite imagery captured during favourable sea conditions, which can vary spatially and temporally. The physical footprint of aquaculture, expected to increase in the next years, is one of the most extensive in the world (Bugnot et al. 2021), with projections for its expansion also in Brazil (FAO 2020). Thus, fishing and aquaculture are important drivers of coastal hardening on less developed areas of SP and may continue to do so (Valenti et al. 2021). The increase of aquaculture, however, may intensify current tensions with local artisanal fisheries (Prado et al. 2022). A better understanding of the extension and spatialization of these activities is crucial for improving spatial marine planning and mitigating those conflicts.

Jetties were the most common coastal infrastructure among ASe, adding up over 40 km of hard structures on the SP coast. These structures were also the second most common type identified in our study, ranking third in all three regions. Pontoons, on the other hand, contributed to an extra 18 km of AS in the study area. Most of the jetties (67%; 369 out of 548) were privately-owned structures associated with houses and marinas. This trend was more pronounced for pontoons (94%; 423 out of 451). The proliferation of jetties and pontoons seems associated with the secondary usage of coastal cities, with fluctuating migrations driven by second-home ownership and tourism in the coast of Brazil (Moraes 2007). These migrations have created and expanded markets for nautical and leisure activities, which encompasses services such as boat sales, rentals, and repairs, marina operation, tourist cruises, boat tours, diving, and offshore charter fishing (da Fonseca et al. 2015; Sanguinet and Sass 2022). Collectively, these activities have increased the demand for building infrastructure for sea accessibility and docking. The remaining jetties are primarily associated with public transportation within and among municipalities, in the form of bridges, ferry and passenger terminals, and public piers.

Amid economic expansion, population increases, and migratory patterns fuelling urbanisation in the Brazilian coast, with a legacy of colonialism and persistent social inequalities, a segment of the population remains outside the formal economy (Moraes 2007). Such marginalised people end up settling in vulnerable territories and protected areas such as floodable terrains, hillslopes, and mangroves (Moraes 2007). In our study we detected that 91% (~10 km) of the total AS extent classified as buildings refers to stilt houses – locally known as palafitas – situated along estuarine margins in the central region of the SP coast. These are extremely precarious habitations sustained by pillars built over wetlands (Fonseca Feitosa et al. 2021). Although substantial, the extent of AS provided by such habitations is underestimated in our study. These settlements spread over intertidal areas once occupied by mangroves and tidal flats, but here we only quantified their linear length. Palafitas in Santos municipality alone, for instance, span over 350,000 m2, while São Vicente, Cubatão, and Guarujá also have these settlements (Supplementary Material, Fig. A5 and A8). Beyond severe humanitarian, socioeconomic, and other environmental problems, these settlements also contribute to coastal hardening in SP – although this is hardly the main concern in this situation. Currently there is a project to transform these habitations, providing better living conditions for people and increasing surrounding environmental quality (Fassina et al. n.d.). This project provides an opportunity for developing a habitational model also taking nature into account in terms of recovering native biodiversity and ecosystem services.

Shifts in physical conditions, habitat loss, and decline in biodiversity and ecosystem functions are some of the main impacts associated with coastal artificial infrastructure (Bulleri and Chapman 2010; Dugan et al. 2012; Heery et al. 2017). In Brazil, a recent review revealed that research on the ecological impacts of coastal hardening has focused on artificial reefs and fishery ecology, with 66 invasive exotic species found in association with AS (A.S. Martinez, pers. comm.). Despite being a major source of environmental impacts, as suggested by our mappings and literature, coastal hardening is still an overlooked theme in national research. Our assessment thereby helps fulfilling an important gap to comprehend the impacts of coastal urbanisation and support management strategies. For instance, data on the extent and localization of AS are essential for understanding potential paths for bioinvasion (Bishop et al. 2017). In the future, expanding mappings of AS and identifying drivers of coastal hardening in Brazil will provide a more thorough comprehension of its impacts and how to manage them. Finally, we highlight that some common and often abundant AS such as artificial reefs, shipwrecks, submarine cables and pipelines, vessel hulls, and offshore platforms (Bugnot et al. 2021; Lemasson et al. 2024) were not mapped in our research due to methodological approach and study scope. The extent of AS in SP coast is, therefore, even larger.

4.2 Occupations in LECZ100m in SP

One-quarter of the soft coastline of SP has inland settlements and urban assets in LECZ100m, mostly (78%) adjacent to sandy beaches. This figure represents more than half of the extent of sandy beaches, reaching 74 and 81% in the north and central regions, respectively (Table 2). These alarming results reveal a process of coastal squeeze far intense than the global pattern, where 33% of sandy shores harbour less than 100 m of infrastructure-free space (Lansu et al. 2024). Extensive occupation of areas near sandy beaches reflects strong real estate speculation upon these highly valued territories, much of it driven by second-home market (Supplementary Material, Fig. A11). Six out of sixteen municipalities on the SP coast (Ilha Comprida, Peruíbe, Itanhaém, Mongaguá, Bertioga, and Ubatuba) has more permanently and occasionally empty houses (52.7 to 62.7%) than permanently occupied ones (IBGE 2022). At the same time, SP has been increasingly facing flooding and coastal erosion due to extreme rains, storm surges, wave impact, and sea-level rise (e.g., de Mahiques et al. 2016; dos Santos and Serrao-Neumann 2018; Nunes et al. 2019). As a result, engineering interventions to protect urban assets have been, and are expected to be, implemented as the impacts of climate change intensify. Beach nourishment, for instance, has recently become popular in Brazil, with some projects already executed (Barros et al. 2021; Soares-Gomes et al. 2023) and many others are planned and underway. Such interventions, however, are a palliative solution that can result in several unwanted impacts (de Schipper et al. 2021). Wherever possible, nature-based approaches, such as conservation and restauration of habitats, deliver long-term better solutions (Morris et al. 2018; Manes et al. 2023). Although promising, such initiatives are not yet a common practice in Brazil (Barros et al. 2021; Soares-Gomes et al. 2023).

Occupations in lowlands near to estuarine and river margins are also expressive in SP, considering that much of the hardened coastline is situated within estuaries. Unlike beachfront locations, these regions have historically been overlooked by the real estate market and are considered permanent preservation areas according to federal environmental law (Moraes 2007). Nonetheless, such territories became main residential areas for low-income and marginalised population, especially in the central region of the SP coast (Fonseca Feitosa et al. 2021). To make matters worse, amendments to the environmental law in 2012 paved the way for legitimising occupation of these territories under the guise of public utility and social interest principles (Azevedo and Oliveira 2014). In the face of sea-level rise and increase in the frequency and intensity of climate-induced coastal hazards, this reality poses major governance challenges due to the overlap of the greatest social and environmental vulnerabilities (Fassina et al. n.d.). In this context, managerial actions towards mitigation and adaptation are urgent to attain climate resilience and climate justice (UN 2019a, 2022b).

5 Conclusion

This study presents a detailed large-scale assessment of coastline urbanisation in the state of São Paulo, providing valuable data for decision-making on urban and marine spatial planning. Our results indicate significant coastal hardening and occupation of lowland areas in São Paulo, with heterogenous patterns along the assessed regions. With increasing coastal hazards, engineering interventions to protect coastlines and restore habitats are expected to intensify. This calls for scientifically sound managerial actions towards mitigating impacts and adapting cities in line with sustainability principles. Without disregarding context-dependent realities, we propose the following strategies:

  1. 1)

    Protection of natural areas. Mangrove fringes, riparian vegetation, and beachfront vegetation should be protected as they provide natural defence against climate-induced coastal hazards, reducing economic and social losses (e.g., Zamboni et al. 2022). Despite the significant presence of preservation areas along the São Paulo coast, only a small portion encompasses immediate beachfront areas. To address this strategy, expansion of existing preservation areas and creation of new ones may be necessary. Furthermore, legal provisions enabling the occupation of permanent preservation areas (e.g., mangroves and river margins) should be reconsidered.

  2. 2)

    Restoration of impacted natural areas. Where feasible, estuarine/river margins and beachfront areas should be revegetated. Nature-based approaches offer promising solutions, with the possibility of implementing 'soft' and hybrid interventions (Morris et al. 2018, 2019; Firth et al. 2024). However, it is important to note that even these methods are not universally applicable or necessarily sustainable practices (Parkinson and Ogurcak 2018).

  3. 3)

    Revitalisation of highly urbanised areas and hard eco-engineering. Upgrading areas already replaced or altered by artificial infrastructure presents an opportunity to manage coastal hazards and restore native biodiversity while sustaining ecosystem services (Mayer-Pinto et al. 2017; Morris et al. 2018, 2019; Airoldi et al. 2021; Firth et al. 2024). Marine eco-engineering offers a viable option to achieve these goals, which depends on local knowledge (Strain et al. 2018). This is also applicable for future interventions in areas where ‘soft’ approaches are not viable (Morris et al. 2018, 2019). This approach differs from traditional engineering by explicitly addressing concern on environmental sustainability. To address this strategy in Brazil, developing research on marine eco-engineering is necessary. In applying this approach caution is needed to avoid bluewashing (Firth et al. 2020, 2024).

Our recommendations aim to promote resilience and sustainability in coastal areas, safeguarding both the environment, the economy, and the well-being of local communities. Their success, however, will inevitably depend on political and social engagement and goodwill. Tackling this challenge requires educational efforts and enhanced dialogue with society to raise consciousness about the perils arising from climate change, as well as the consequences of unsustainable coastal development and exploitation.