The role of urbanization in facilitating the introduction and establishment of non-native animal species: a comprehensive review

Abstract

While urbanization is often associated to a loss of biodiversity, non-native animal species are strikingly successful in urban landscapes. As biological invasions are recognized to have detrimental environmental, social and economic impacts, extensive understanding of the interactions between invasive species and the abiotic and biotic environment is necessary for effective prevention and management strategies. However, the mechanisms underlying the success of invasive animals in urban environments are still poorly understood. We provide a first conceptual review of the role of urbanization in the introduction, establishment, and potential spread of non-native animal species. We summarize and discuss the mechanisms enhancing biological invasive potential of non-native animals in urban environments, by both isolating and interlinking the abiotic and biotic drivers involved. Following the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA 2020) process, this systematic review covers a total of 124 studies comprehensive of all taxonomic groups, albeit with an evident publication bias for avian and terrestrial invertebrate species (22.1% and 19.8% of literature respectively). High-income regions also represent a larger bulk of the literature (Europe: 26.7%, North America: 23.7%). The most common reported factors facilitating species invasions in urban areas are reduced biotic resistance, and the competitive and urban-compatible ecological and/or behavioural traits of non-native animals allowing urban exploitation and aiding invasion. Finally, we identify important knowledge gaps, such as the scarcity of studies investigating socio-economic spatial patterns in the presence and abundance of invasive species, as well as the adaptive evolution of non-native animal species in urban areas.

Introduction

Invasive non-native animal species are non-native animal species (NNAS) that have established a large self-sustaining population, maintained over multiple generations and dispersed away from its point of introduction (Colléony and Shwartz 2020, Santana Marques et al. 2020). Previous studies have suggested that NNAS can be strikingly successful in urban areas (Hima et al. 2019, Santana Marques et al. 2020, Rogers et al. 2021). On one hand, cities can supply constant resource availability throughout the year and predation is often relaxed in urban areas (Carthey and Banks 2018, García-Arroyo et al. 2020, Santana Marques et al. 2020). On the other hand, cities are also associated with a range of pollutants, such as light at night, noise, chemicals, heavy-metals, increased ambient temperatures, habitat loss and fragmentation, and novel impervious surfaces and structures (Rawson et al. 2010, Harrison and Winfree 2015, McMahon et al. 2017, Buchholz and Kowarik, 2019, Santana Marques et al. 2020). Whilst native species are consequentially negatively associated with urban land cover (Riley et al. 2005, Blair and Johnson 2008, Botham et al. 2009, Isaac et al. 2014, La Sorte et al. 2018, Luo et al. 2018), both NNAS richness and abundance are positively correlated to increasing urbanization, especially in invertebrates (Riley et al. 2005, Blair and Johnson 2008, Cadotte et al. 2017, Rogers et al. 2021). In fact, many of the most common urban species are invasive (Shochat et al. 2010, Bonnington et al. 2014).

Invasive NNAS alter ecosystem function and native community structure, promoting simplification of communities and biotic homogenization (Grarock et al. 2014, Mbenoun Masse et al. 2017, Hima et al. 2019, Graf et al. 2020, Barsotti et al. 2021). Native species are reduced or displaced through competition for resources and breeding sites, predation and hybridization (Short and Petren 2012, Tennessen et al., 2016, Sicuro et al. 2017, Broughton 2020, Colléony and Shwartz 2020, Fibla et al. 2020, García-Arroyo et al. 2020, Barsotti et al. 2021, Lamb et al. 2021, Taggar et al. 2021). Additionally, NNAS may carry harmful pathogens and parasites (Haag-Wackernagel and Moch 2004, Ficetola et al. 2007, Vázquez and Sánchez 2015, Miller et al. 2017, Mori et al. 2018a, 2018b, Hernández-Brito et al. 2020, Sandoval-Rodríguez et al. 2021), and can be serious agricultural pests (Reed et al. 2013, Sicuro et al. 2017, Cesari et al. 2018, García-Arroyo et al. 2020, Godefroid et al. 2020, Diagne et al. 2021). Some NNAS may also cause damage to infrastructure (García-Arroyo et al. 2020, Diagne et al. 2021) and provoke human and animal allergic reactions (Godefroid et al. 2020). Knowledge on the extent of the impacts of some of the globally most common invasive NNAS are still lacking (García-Arroyo et al. 2020).

Because of the widely reported effects of invasive NNAS, biological invasions are now internationally recognized to have detrimental environmental, social and economic impacts (Roy et al. 2014, Parejo et al. 2015, Padayachee et al. 2017, Colléony and Shwartz 2020, García-Arroyo et al. 2020, Gethöffe and Siebert 2020), and are listed as the second major driver of species extinction (Parejo et al. 2015, García-Arroyo et al. 2020, Tricarico et al. 2021), particularly for vertebrate populations (Pollock et al. 2019, Barsotti et al. 2021). Still, biotic range shifts have been happening at unprecedented rates world-wide for some decades (Godefroid et al. 2020). Prevention is the most cost effective and environmentally desirable defence strategy against NNAS (Wittenberg and Cock 2001, Roy et al. 2014, Gethöffe and Siebert 2020, Graf et al. 2020). Effective prevention relies on understanding biological invasions and the interactions between NNAS and the abiotic and biotic environment (Sato et al. 2010, Suppo et al. 2018). Yet, the mechanisms underlying the success of NNAS in urban environments are still poorly understood (Urban et al. 2008, Johnston et al. 2017, Santana Marques et al. 2020, Vanek et al. 2021). It is difficult to isolate mechanisms influencing invasion potential, as these are often confounded with different conditions influencing vulnerability to invasions (Crooks et al. 2011).

This review aims at summarizing and explaining the mechanisms enhancing biological invasive potential in urban environments; and ultimately the role of urbanization in the establishment and spread of NNAS. After a systematic literature search, a total of 124 studies were selected and reviewed, building a comprehensive collection of knowledge involving the introduction pathways of NNAS and urbanization, and the abiotic and biotic urban drivers implicated in the facilitation of establishment of invasive species. Moreover, the characteristics of urban animals leading to their success in anthropogenic landscapes were taken into consideration to better understand the success of NNAS in urban environments.

Materials and methods

Literature review

A first literature search of the Web of Science Core Collection (WoS) and Scopus search engine was performed in December 2020. The search aimed to provide comprehensive literature on invasive animal species in urban environments, aiding in identifying major urbanization drivers involved in the establishment and spread of NNAS. We conducted this search using the keywords (KW): (1) KW=(‘urban’ AND ‘biological invasive potential’) and (2) KW=(‘urban’ AND ‘invasive species’). 22 papers were thus obtained, including comprehensive reviews on the topic as well as focused case study papers (Table 1).

Once drivers involved were identified (Table 2), the research was continued (January–March 2021) on the same databases (databases time-frame covered: 1990-03.2021; date last checked: 29.02.2024): (3) KW=(‘urban’ AND ‘invasive species’ AND (‘introduction’ OR ‘pathway’ OR ‘exploiter’ OR ‘breeding’ OR ‘nesting’ OR ‘resources’ OR ‘resource availability’ OR ‘pollution’ OR ‘climate’ OR ‘homogenisation’ OR ‘biotic resistance’ OR ‘predation’ OR ‘pathway’ OR ‘vegetation’ OR ‘water availability’ OR ‘pollination’ OR ‘spread’)), resulting in 1,663 results on WoS, 45 on Scopus; and (4) KW=(‘urban’ AND ‘animal assemblages’), resulting in 360 results on WoS, 425 on Scopus. Further 28 useful papers were identified through citation searching from papers obtained from searches (1), (2), (3) and (4); if referenced papers were not contained in the Scopus or WoS databases, they were obtained through the University of Glasgow Library digital search engine. To comprehensively include literature across taxa, a further search was performed, in January–March 2021: (5) KW=(‘urban’ AND ‘invasive species’ AND (‘bird’ OR ‘avian’ OR ‘reptile’ OR ‘amphibian’ OR ‘invertebrate’ OR ‘marine’ OR ‘mammal’ OR ‘rodent’)), producing 466 more results on WoS, 521 on Scopus. All produced records were collated into a an excel spreadsheet and processed for de-duplication. After de-duplication, 2583 unique records remained. For a detailed schematic overview of the literature search, see the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 Flow Diagram in Fig. S1 of Supplementary Materials 2 (Page et al., 2021). The full data-set of records obtained through the PRISMA 2020 procedure, at each selection stage, is available in spreadsheet form of Supplementary Materials 3.

On 27 February 2024, 4 additional papers were retrieved from WoS and Scopus, to further strengthen specific sections of the review. These papers were identified through recommendations from experts.

Criteria for inclusion

The interest is in exploring the role of urbanization in facilitating the introduction and establishment of NNAS. After careful reviewing the 22 papers (see Table 1) from searches (1) and (2), a set of abiotic and biotic drivers affecting non-native animal species introduction and establishment were identified (see Table 2). The 2583 unique records produced by searches (2), (3), (4) and (5) were screened by the corresponding author to be selected for if relevant and constructive to the listed drivers and other informative discussion topics (Table 2). Screened records were also excluded if discussing:

• plant invasive alien species only,

• invasive species in non-urban environments or non-native to native species interactions without providing sufficient ulterior knowledge significantly aimed towards the focused review topic,

• reports on protozoan parasites,

• urban environment characteristics and disturbances not deemed significantly useful towards the review’s objectives.

Finally, 353 studies were identified and retrieved. See Fig. S1 for PRISMA 2020 flow diagram for detailed schematic overview of criteria for inclusion process, available of Supplementary Materials 2.

Study risk of bias assessment

Screening a selection of studies was carried out by the corresponding author independently. No automation tools was used. A study was included if relevant to the review topic and pertinent to the inclusion criteria.

Synthesis methods

All studies finally selected for inclusion were carefully read and summarized, with key arguments noted or extracted. The arguments were then re-organized and grouped for synthesis within subheadings rapresentant of the specific urban abiotic and biotic drivers to be discussed; information was also compiled for other relevant discussion topics (Table 2).

Results

Study selection

Out of the retrieved studies, 124 studies were ultimately included in the review. Records were excluded if not meeting the described inclusion criteria, but accidentally retrieved. Studies were also excluded if deemed redundant (i.e. already a sufficient number of studies and exemplar case studies found per mechanism) or not significantly valuable at furthering discussion of identified discussion topics, or of newly emerged topics, based on the literature already confirmed for effective inclusion throughout the selection process. This review aims at discussing concepts using exemplar case studies with a good representation of described taxa and geographic location of literature. Therefore, an effort was made to include a selection of literature reflecting these aims.

The full list of resulting literature reviewed (n = 124) can be found in Table S1 of Supplementary Materials 1, summarized and organized by urban drivers (abiotic and biotic), with specified taxon in focus and world region where the studies were performed, as well as further details on the relevant drivers discussed in each study.

Study characteristics

The literature reviewed spans a wide range of taxonomic groups, although avian and terrestrial invertebrate arthropod species receive more far attention (22.1% and 19.8% of reviewed literature respectively), and aquatic species and herpetofauna are understudied (see Fig. 1). Studies performed in Europe and North America composed the largest bulk of literature (26.7% and 23.7% respectively), with remarkably low representation from the Global South. Africa and Asia are represented by only 4.6% each; the Middle East (1.4%) is lacking particularly behind. Studies taking place in Panama were considered as North or South America if data-collection occurred North or South of the Panama Channel respectively. The most common reported factors facilitating species invasions in urban areas are impoverished community structure and biotic homogenization, and the competitive and urban-compatible ecological and/or behavioural traits of non-native animals allowing urban exploitation and aiding invasion (Table 2). The least reported urban driver is within-habitat heterogeneity and the forest-edge effect, discussed in only 4 of the reviewed studies.

The introduction and dispersal of non-native animal species

A large proportion of NNAS introduction is unintentional (Suppo et al. 2018), mostly happening through commercial hubs at transport networks (ie seaports, airports, road and river transport networks) (Hima et al. 2019, Godefroid et al. 2020). Travel and tourism are also to be accounted for (Padayachee et al. 2017). This is referred to as the ‘stowaway’ pathway (Padayachee et al. 2017). Insects (especially social insects) are the most popular invaders world-wide; their small size easily grants them accidental transportation, and r-selection life-strategy allows for high propagule pressure (ie quantification of the organisms introduced into the community and number of introduction events) (Banha et al. 2017, Suppo et al. 2018). Small mammals (e.g. rodents) are also typically introduced as stowaways (Anderson 2009). Otherwise, invertebrates are often introduced through the ‘contaminant’ pathway (Padayachee et al. 2017), where NNAS are transported unintentionally together with an intentionally transported commodity (e.g. commensal species, food, plants, etc). The contaminant pathway differs from the stowaway pathway by being predictable from knowledge of the shipped commodities and their shipping routes (Hulme et al. 2008). Non-native ornamental vegetation for urban landscaping and private gardens can often carry highly invasive and undesired species (Bella 2013, Čeplová et al. 2017, Godefroid et al. 2020), such as the oak processionary moth (Thaumetopoea processionea) (Godefroid et al. 2020). Horticultural activities and transportation of timber or fresh/live comestibles transportation may also introduce exotic pests (Marzluff 2005, Hulme et al. 2008).

Most vertebrate NNAS are introduced intentionally, for landscape and fauna improvement or above all through pet trade, to pet shops and homes (Padayachee et al. 2017, Maceda-Veiga et al. 2019, Hernández-Brito et al. 2020). Spread is by escape and unaided dispersal (Padayachee et al., 2017), or release (Hulme et al. 2008). The exotic pet trade is key for avian invasions (García-Arroyo et al. 2020), and pet release is considered the most important pathway for reptiles, such as the highly invasive red-eared slider turtle (Banha et al. 2017) or Burmese python (Orzechowski et al. 2019).

Disturbance from agricultural intensification and urbanization causes native fauna to decline (Marzluff 2005, Colléony and Shwartz 2020), particularly specialist and susceptible species (Padayachee et al. 2017, Colléony and Shwartz 2020). Contrastingly, NNAS propagule pressure in urban areas can be relatively high (Banha et al. 2017; Cadotte et al. 2017). With over 55% of the human population now living within urban areas (Ritchie and Roser 2018), cities are hubs of movement of people and goods (Padayachee et al. 2017). Thus, cities are often the first and most prominent introduction point (Cadotte et al. 2017, Padayachee et al. 2017, Hernández-Brito et al. 2020).

Urban settlements are interconnected by transport corridors, both aquatic (ie rivers and canals) and terrestrial (ie roads) (Botham et al. 2009), facilitating the natural and un-aided dispersal of NNAS to the surrounding areas and other urban centres. NNAS can easily travel through their preferred urban and peri-urban conditioned environments avoiding the more difficult and unfitting native natural conditions. For instance, the Asian tiger mosquito (Aedes albopictus) has successfully reached continental and global expansion by both dispersing naturally alongside roads and long-distance human-aided dispersal (Sherpa et al. 2020); as stowaway in passenger cars (Eritja et al. 2017) and through importation of goods, such as used tires. A.albopictus is a disease vector for dengue and dengue haemorrhagic fever (Knudsen et al. 1996). Inter-city roads, much like cities themselves, may also provide easily accessible food. Thus, promoting the dispersion of synanthropic NNAS, as already documented for several invasive birds (D’Amico et al. 2013).

What makes an urban animal: a trait-based approach

Recognition of the basic characteristic of the urban environment, and what makes a potential urban animal, is crucial to understanding invasive success on NNAS in urban environments (Kark et al. 2007, Suppo et al. 2018). Acting as a filter for native biodiversity (Alberti 2015, García-Arroyo et al. 2020), the urban environment selects against more specialized and susceptible fauna (Marzluff, 2005, Isaac et al. 2014, Padayachee et al. 2017; Colléony and Shwartz 2020). A process reminiscent of invasion biology, where a few species take over the habitat and dominate in abundance (Kark et al. 2007, Colléony and Shwartz 2020). Urban fauna can be placed under three categories: urban avoiders, exploiters, and adapters (Kark et al. 2007, Rogers et al. 2021). Urban avoiders are native species that cannot occur in disturbed city areas as incapable of adapting to the urban environment (Kark et al. 2007, Colléony and Shwartz 2020). Urban adapters capitalize on urban resources (Colléony and Shwartz 2020) and thrive at intermediate levels of urbanization (e.g. suburbs); they can be both native and invasive (Kark et al. 2007). Where urbanization is most intense, urban exploiters reach high densities (Gering and Blair 1999, Colléony and Shwartz 2020). Exploiters can benefit from urbanization, by adapting to exploit the urban environment (Colléony and Shwartz 2020) and forming commensal relationships with humans, frequently becoming dependent on urbanization and ubiquitous globally (Kark et al. 2007). Urban exploiters are most often native synanthropic species or exotic (Shochat et al. 2010; Colléony and Shwartz 2020).

Urban exploiters have advantageous life-history traits that respond positively to the ‘urban filter’ (Gering and Blair 1999, Santana Marques et al. 2020). Sociality amongst conspecifics improves foraging ability (improved localization and communication of food sources) (Kark et al. 2007, Shochat et al. 2010), avoidance of predators, and eases competition as well as potentially boosting boldness and growth; sociality amongst heterospecifics allows for the same benefits of belonging to a larger group when availability of conspecifics is low, which may be particularly relevant for NNAS at the first invasion stages (Kark et al. 2007, Camacho-Cervantes et al. 2023). Urban species are also commonly sedentary, as sedentary avian species in a relatively stable urban environment are likely to monopolize the usually scarce nesting sites available in cities whilst migrants are away (Kark et al. 2007). In fact, an animal’s breeding ecology is in itself a key factor for urban success, with reproductive life-histories, breeding site and ‘nest’ structure and reproductive output having to reflect the selective demands of novel city ecosystems (Gering and Blair 1999, Kark et al. 2007). Diet plays an important role in urban success (Kark et al. 2007). Food resources can be increased under urbanization, but are less varied and often novel (Carthey and Banks 2018, Colléony and Shwartz 2020). For bird species, granivores, fructivores and especially omnivores (Jerusalem: 50% of urban-downtown avian species) are favoured, whilst invertebrate feeders decrease; ground invertebrate-feeders persist (Kark et al. 2007). Urban successful species are usually ecological generalists (Colléony and Shwartz 2020) with affinity for feeding innovations and adapting to new foods, such as scavenging on human refuse. Behavioural flexibility (Kark et al. 2007) and dominant or aggressive interspecific behaviours allow for effective use of resources and upper-hand in competition for food and breeding patches (Ficetola et al. 2007, Shochat et al. 2010).

Urban-successful NNAS are commonly urban exploiters (Kark et al. 2007, Shochat et al. 2010), and typically possess the life-history traits desired for invasive success (Gering and Blair 1999). Human-commensal and intentionally human-introduced species may have the further advantage of pre-adaptation to human-modified environments (Kark et al. 2007, Zozaya et al. 2015). Feral pigeons, once domesticated and kept in cities also outside of their native range for food, as homing pigeons or as pets, are now invasive in cities across the globe (Konijnenberg and Goerlich-Jansson 2020). House geckos living as human commensals have adapted to exploit the novel ‘artificial light’ niche, aiding colonization of urban environments in non-native regions, and improving competition for food against native less adapted geckos (Zozaya et al. 2015). Highly invasive house sparrows and rats have also been associated with humans for millenniums (Anderson 2009, García-Arroyo et al. 2020). In terms of advantageous traits, house sparrows are ecologically and physically plastic, and aggressive towards both conspecifics and heterospecifics (García-Arroyo et al. 2020). Rats are generalist omnivores that, through high evolutionary potential by high fecundity and rapid generations, have achieved immunity to poisons (Collins et al. 2000, Anderson 2009). NNAS may also act as vectors of parasites (Hernández-Brito et al. 2020) and pathogens from their native range (e.g. American bullfrog and chytridiomycosis (Ficetola et al. 2007)). Carrying disease may be advantageous towards invasive-success when natives occupying their niche are vulnerable to the same disease: e.g. asymptomatic invasive American eastern grey squirrel widely displace native Eurasian red squirrel by fatal Squirrel Pox Virus infections (Chantrey et al. 2014).

The urban habitat as facilitator of NNAS

Urban landscapes are uniquely altered habitats (Collins et al. 2000, Isaac et al. 2014, Cadotte et al. 2017), where abiotic and biotic conditions are artificially maintained, contrasting to the surrounding natural environment (Alberti 2015). Deep modifications are produced on the habitat structure, local seasonal variation and temporal variability of resources, water systems as well as climate and temperature (Collins et al. 2000, Alberti 2015, Rooney et al. 2015, Richmond et al. 2018). The structural and biotic conditions also prompt changes in soil structure and nutrient dynamics of the ecosystem; whilst some nutrients are mobilized (e.g. phosphorus and nitrogen pollution nutrient pollution), others are depleted (Collins et al. 2000, US EPA 2013, Cadotte et al. 2017). Green spaces are reduced and patchy, replaced by novel anthropogenic structures and non-permeable surfaces (Groffman et al. 2014, Buchholz and Kowarik 2019, Colléony and Shwartz 2020). These remaining green spaces lack tree cover, increasing the amount of sun-light to the ground and thus promoting growth of weedy vegetation (Cadotte et al. 2017). Vegetated ground cover is overall largely simplified, often purposely maintained as homogenous short grass (Marzluff 2005; Groffman et al. 2014; Isaac et al. 2014; Peng et al. 2020; Rogers et al. 2021). Native vegetation is scarce in cities (Marzluff 2005, Kark et al. 2007), with exotic vegetation being intentionally introduced as ornamental plants or for horticulture (Hulme et al. 2008, Groffman et al. 2014, Godefroid et al. 2020). In the UK, plants introduced after the 1500 dominate urban assemblages, whilst archeophytes are no longer strongly associated to urban land-cover (Botham et al. 2009).

Such unique and novel habitat characteristics are hostile for many, but become beneficial to NNAS that can tolerate the unconventional conditions and capitalize on the many offered opportunities (Collins et al. 2000; Isaac et al. 2014; Cadotte et al. 2017; Luo et al. 2018). Moreover, increased forest edge effect by fragmentation of natural habitats and high-within habitat heterogeneity, combined with novel anthropogenic disturbances maintaining habitats at an early successional stage, suggests a fast generation of unique and colonizable niches (Marzluff 2005, Alberti 2015, Padayachee et al. 2017). Examining the interactions between the abiotic environment and NNAS is important for understanding biological invasions (Urban et al. 2008, Sato et al. 2010, Suppo et al. 2018).

Climate and urban heat island effect

The most important environmental factor facilitating biological invasions is climate (Ficetola et al. 2007, Vimercati et al. 2018). Invasion potentials are often predicted based on the affinity of the regional climate and the realized niche of the targeted NNAS in its native range (Ficetola et al. 2007). Urbanization often ameliorates the local climate (Collins et al. 2000, Cordonnier et al. 2020, Hernández-Brito et al. 2020), opening new niches that NNAS can exploit. The most influential climatic factor is temperature, particularly for ectotherms and species with temperature-dependant sex determination (Ficetola et al. 2007, Banha et al. 2017, Javal et al. 2018). Due to decades of human activities, such as burning of fossil fuels, and the reduction of photosynthetic organisms (Collins et al. 2000, US EPA 2015, Santana Marques et al. 2020), carbon dioxide concentrations in cities are far higher than the global average. This causes temperatures within cities to significantly exceed those of the surrounding areas, a process known as the ‘urban heat island’ effect (Collins et al. 2000, Alberti 2015). The heterotrophic nature of urban systems is also an important factor: a typical city is estimated to convert into heat 70 times more usable energy per square meter than its close natural surroundings (Collins et al. 2000). Precipitation, another important climatic feature (Ficetola et al. 2007), is also impacted by anthropogenic activities. Regional precipitation patterns are affected by the regular fluctuations of weekly cycles of air pollution, air carbon-monoxide and ozone levels accumulating during the working weekdays, leading to a higher probability of precipitation during weekends (Cerveny and Balling 1998, Collins et al. 2000). The milder microclimate of cities extends growing seasons in temperate regions, while irrigation in arid tropical and subtropical regions prevents extended droughts (Alberti 2015). These conditions favour tropical non-indigenous animal and plant species (Alberti 2015, Cadotte et al. 2017, Hernández-Brito et al. 2020). Human utilities also become involved in ameliorating tolerance to extreme climates (Cordonnier et al. 2020, Mills and McGraw 2021). For instance, rosy-faced lovebirds in Phoenix (USA) use relief-air vents on building surfaces to cool down in summer, and European starlings perch atop of chimneys in lower winter temperatures (Mills and McGraw 2021).

Novel structures and surfaces

Urbanization gives rise to novel structures, convenient to humans but composed of impractical, and often also rough and unpleasant materials; such as concrete, metal or plastic/glass (Gering and Blair 1999, Groffman et al. 2014). Regardless, anthropogenic structures provide excellent breeding and roosting/nesting opportunities for many NNAS (Kark et al. 2007, Cordonnier et al. 2020, Arsenault-Benoit et al. 2021). Dispersal and breeding of invasive mosquitos is largely dependent on anthropogenic conditions; demanding standing water-filled artificial containers (ie used tired, flowerpots, abandoned bottles and jars) (Sherpa et al. 2020) to deposit their eggs and mature to adulthood (Eckhoff 2011). Moreover, the degeneracy of the habitat allows for high predictability and visibility of food for urban exploiter generalists (Anderies et al. 2007, Carthey and Banks 2018). Increased open spaces and absence of tree cover are particularly beneficial for air mobile organisms. Urban landscapes and contained structures are increasingly similar across different human settlements, advantaging species pre-adapted to human structures in invading other regions containing humans (Kark et al. 2007, Crooks et al. 2011, Zozaya et al. 2015). This is usually the case for exotics introduced via the pet trade, as they have been adapted or exposed to anthropogenic conditions (Crooks et al. 2011).

Availability of breeding sites

The variety of breeding locations (e.g. nest/den opportunities, fish and amphibian spawning grounds) is skewed in urban environments (Cadotte et al. 2017). Native vegetation is scarce, dead trees are removed (Kark et al. 2007, Diamond and Ross 2019), and wetland or water systems are often absent or largely diminished and altered (Groffman et al. 2014). Ground nesting is also disfavoured by the vast amount if impervious surface, and the portion of ground unused by humans is threatened by predatory domestic animals (ie cats and dogs) (Kark et al. 2007). Hence, the ability to exploit novel breeding sites is a key factor for urban success (Gering and Blair 1999, Kark et al., 2007, Lermite et al. 2021). Black rats are highly flexible nesters; nesting in tree cavities, on the ground and inside buildings (Dowding and Murphy 1994, Cox et al. 2000). European starlings, listed within the 100 most invasive species throughout the world, also show great flexibility in roost-site selection, allowing for eased colonization of non-native urban sites through transition to the available colonizable urban landscape resources (Clergeau and Quenot 2007). Buildings translate to cliff-sides or rock, so avian cliff, cavity and rock nesters are advantaged. For instance, feral pigeons see cities as habitats rich in nesting availability as they descend from rock doves (Columba livia), which naturally nest at seaside cliffs. With their native predators removed, they have become one of the most successful and cosmopolitan bird species, having invaded cities in every continent (excluding Antarctica) (Konijnenberg and Goerlich-Jansson 2020). Another highly successful invasive urban bird is the house sparrow, as their aggressive behaviour aids effective nesting when competition for resources is high (García-Arroyo et al. 2020). Monk parakeets (Myopsitta monachus; NNAS in North America and Western Europe, Asia, Africa and some oceanic islands) instead avoid competition for nest sites altogether, by building their own nests (Hernández-Brito et al. 2020). Canopy nesters are also common in cities and suburbs (Kark et al. 2007). Furthermore, exotics make use of introduced non-native vegetation they may be pre-adapted to for roosting and breeding (Gering and Blair 1999, Shochat et al. 2010, Godefroid et al. 2020).

Habitat homogenization, aquatic features and water availability

Urban habitats are designed to best fit human activities (McKinney 2006, Alberti 2015). Regardless of within-habitat heterogeneity relying partly on the choice of individual home owners (Braschler et al. 2020, Colléony and Shwartz 2020), different anthropogenic habitats across the globe are still highly similar. More similar to each other than their respective natural environment (McKinney 2006, Alberti 2015). Recurring structural patterns such as altered plant ecology by replacement of natural vegetation assemblages by turfgrass, popular and horticultural plant species, as well as the addition of impervious surfaces (e.g. asphalt) alter the local soil ecology via changes in soil moisture and organic matter. In naturally arid environments soil organic matter is greatly increased and in humid environments it is slightly decreased. Other structural aspects of the urban environment that can have ecological impacts are roads and residential landscape planning (Groffman et al. 2014). Continental and global homogenization reduces extreme and unfavourable environments, improving net ecosystem compatibility for NNAS (Ficetola et al. 2007, Vimercati et al. 2018, Barsotti et al. 2021).

Water availability and aquatic features of an ecosystem are limiting factors for many species (Gregg et al. 2019, Barsotti et al. 2021), particularly for aquatic or semi-aquatic organisms needing permanent surface water arrangements for breeding or foraging purposes (Ficetola et al. 2007). Urbanization deeply modifies the structure, distribution and character of the hydrography of natural landscapes. Introduction of new aquatic systems and/or removal or alteration of locally abundant systems leads to a homogenization of habitats towards an intermediate ideal between wet and dry (Collins et al. 2000, Groffman et al. 2014). In temperate or humid zones there is usually a loss of natural channel networks, reduction of lakes and wetlands and redirection of rivers/streams, with simultaneous creation of artificial canals, ponds and reservoirs for flood control, drainage and fill, water supply or recreation. In arid areas surface water systems are introduced for more productive (and aesthetic) land (Groffman et al. 2014, Rooney et al. 2015, Richmond et al. 2018). Moreover, water in naturally arid biomes is made available by artificial ponds and irrigation in gardens, suburban area lawns and agricultural grounds (Alberti 2015, Vimercati et al. 2018); in the USA suburban private lawns were found to make the largest part of all urban land cover (Groffman et al. 2014). The guttural toad (Amietophrynus gutturalis), endemic to more humid regions of southern Africa, have spread to drier African areas. Though they still show physiological responses to the abiotic stressor and dryness, sheltering by human-created water sources (e.g. artificial ponds) and irrigation allows survival and selection to conquer the new environment (Barsotti et al. 2021).

Anthropogenic resources

NNAS are welcomed into cities by an abundance of resources provided by humans (Clergeau and Vergnes 2011, Carthey and Banks 2018, Hernández-Brito et al. 2020); which they can forage for efficiently given the lowered predation pressure (Shochat et al. 2010). Resource availability in urban forests can be 2–3 times higher than in rural forests for generalist avian species, and in city habitats possibly 4 times higher (Shochat et al. 2010). This surplus is credited to both intentionally (e.g. bird feed) and unintentionally (garbage, ornamental and agricultural vegetation) human-provided resources (Clergeau and Vergnes 2011, Galbraith et al. 2015, Colléony and Shwartz 2020). Species in downtown areas rely particularly on human refuse. Diet is a limitation, as certain types of resources are reduced by urbanization. Herbivores, granivores and fructivores prevail over carnivores and insectivores, and generic omnivores are most favoured (Kark et al. 2007, Galbraith et al. 2015). NNAS quickly grow bold and efficient when foraging on urban resources, with little fear of humans and novel objects, but retaining neophobic behaviour towards unfamiliar stimuli. Thus, limiting risk from unknown toxins, while maintaining the need to expend energy exploring new areas and food sources (Candler and Bernal 2015; Carthey and Banks 2018).

The introduction of non-native vegetation is greatly beneficial to NNAS feeding (Gering and Blair 1999, Colléony and Shwartz 2020). Invasive pollinators exploit ornamental flowering plants similar to those of its native range, if not from its native range (Cadotte et al. 2017). Some natural ecosystems are not characterized by abundance of fleshy fruit, but fructivorous NNAS can rely on introduced non-native fruit bearing vegetation (Collins et al. 2000, Cesari et al. 2018). Urbanization can also increase food supply indirectly. Organic pollution as a result of contamination from sewage in urban streams increases abundance of highly nutritious chironomid larvae, a great source of food for invasive fish species (Santana Marques et al. 2020). In Rio de Janeiro, Brazil, invasive guppy fish (P. reticulata) reached 26 times higher densities and larger body sizes in disturbed urban streams than undisturbed rural streams. Physical condition and fecundity were also higher. Resource availability was high enough to negate predation pressure when other fish were present in the stream (Santana Marques et al. 2020).

Chemical and heavy metal pollution

As inorganic pollutants are recognized to have a wide range of adverse effects on exposed organisms, environmental contamination by chemical and heavy metal pollutants from fouling anthropogenic activities raises health concerns for both humans and animals (Saaristo et al. 2018, Camacho-Cervantes and Wong 2023). This is especially true in anthropogenic landscapes, where urbanization concentrates polluting human activities (Almeida et al. 2014, Peng et al. 2020, Camacho-Cervantes and Wong 2023); such as urban/industrial/agricultural runoff, sewage and industrial waste discharge, antifouling and preservative products (Rawson et al. 2010, Crooks et al. 2011, Varó et al. 2015, Roveri et al. 2020). Still, urban polluted systems are often positively associated with NNAS. For example, Copper (common polluting heavy metal, lethal to marine invertebrates in high concentrations) exposure causes declines in abundance and diversity (>40%; Crooks et al. 2011) of natives, but NNAS either show no change or even increase in abundance (Crooks et al. 2011). In fact, urban contaminated aquatic systems are more prone to biological invasions than adjacent open and less contaminated areas (Crooks et al. 2011, McKenzie et al. 2011).

Aquatic chemical and heavy metal pollution in a major driver of biodiversity loss and animal community impoverishment at urban sites (Luo et al. 2018, Peng et al. 2020, Roveri et al. 2020), weeding out sensitive species especially at lower trophic levels with consequential detrimental trophic-cascade effects (Almeida et al. 2014, Saaristo et al. 2018). Moreover, chemical pollution reduces overall fitness of native communities, through direct and indirect effects on physiology, behaviour, and reproductive success. For example, exposure to synthetic chemical agricultural products (i.e. organochlorine pesticides, endosulfan insecticide) reduces parental care behaviour in predatory birds and disrupts pheromonal communication between sexes in newts, whilst round gobies found in chemical and heavy metal contaminated industrial sites show slower mobility and decreased aggressive behaviour (Saaristo et al. 2018). These effects create conditions of decreased biotic resistance and predation pressure for introduced NNAS (Saaristo et al. 2018, Camacho-Cervantes and Wong 2023).

For various organisms, chemical pollution also compromises sensory perception and communication between conspecifics, potentially leading to altered interspecific behaviour between natives and NNAS. This allows for NNAS to take advantage of heterospecific sociality benefits with native species unable to detect them as foreign (Saaristo et al. 2018, Camacho-Cervantes and Wong 2023). Here, heterospecific sociality has been recognized to improve foraging success, provide protection from predators (Camacho-Cervantes et al. 2023), and allow for hybridization for exchange of beneficial alleles with closely related natives, before eventual out-competition (Camacho-Cervantes and Wong 2023).

Tolerance to pollutants is important for the successful exploit contaminated systems. Multiple studies now prove tolerance to chemical and heavy metal pollutants in invasive NNAS (Crooks et al. 2011, McKenzie et al. 2011, Almeida et al. 2014, Saaristo et al. 2018). For instance Watersipora subtorquata, an invasive Bryzoan typically introduced to copper-polluted urban harbour bays and estuaries via hull fouling, was found to possess copper tolerance as a genetically heritable trait (McKenzie et al. 2011). Higher levels of heavy-metal pollution induced production of larger, more pollution-tolerant larvae. Larger larvae can also swim longer, grow faster and reach reproductive maturation earlier, all of which confer advantages in a polluted environment. But large larvae are also energetically expensive to produce, leading to lower fecundity. The inconsistent and fluctuating levels of heavy metal pollution exposure of urban systems selects for plasticity in variability of larvae size production, which has been suggested as an underlying mechanism of increased copper tolerance (McKenzie et al. 2011). Such findings may reflect an anthropogenically-induced selective process that occurs in urban settings. Alternatively, selection of individuals with pre-adaptation to high copper concentration may occur during the transport phase (Crooks et al. 2011).

Tolerance to anthropogenic contaminants has also evolved in terrestrial NNAS resisting pesticides and herbicides (McKenzie et al. 2011). Invasive Artemia franciscana was found to be more tolerant to chlorpyrifos (an organophosphate pesticide) than its native sibling species A. parthenogenetica. Both species showed tolerance to higher ranges of the toxicant, but the fecundity of A. franciscana was less affected. Fecundity is an out-competitive advantage leading to colonization and establishment (Varó et al. 2015). Signs of adaptation to trace lead contamination have also been reported in Australia-invasive house sparrows (Andrew et al. 2019).

Interspecific interactions and NNAS in urban habitats

Community structure and biotic homogenization

A key theme in urban ecology is ecosystem health. Cities are islands of highly simplified and weakly integrated biological communities that are often poor in native (including endemic) species. These factors facilitate biological invasions by opening niche space while relaxing or removing biotic resistance (predation and competitive interactions) (Shochat et al. 2010; Alberti 2015, Cadotte et al. 2017, García-Arroyo et al. 2020). Anthropogenic disturbances weed out sensitive species, and altered and degraded habitat and abiotic conditions filter local assemblages to a selected few (McKinney 2006; Rawson et al. 2010, Alberti 2015, Cadotte et al. 2017, Padayachee et al. 2017, Luo et al. 2018, Newbold et al. 2018). Native bird diversity is generally negatively associated with urbanization, albeit multiple urban bird diversity studies have found overall bird biodiversity (natives and non-natives) to peak at intermediate urban levels (Marzluff 2005, Batáry et al. 2018). On the greater urbanized gradient, loss of green spaces, reduction of native vegetation and overall habitat simplification removes niches, skew breeding opportunities and variety of resources, and compromises life-history strategies; disfavouring specialized and susceptible species (Marzluff 2005, Kark et al. 2007, Cadotte et al. 2017, Colléony and Shwartz 2020, Rogers et al. 2021). Contrastingly, intermediate urbanization may provide favourable conditions of resource abundance and relaxed predation, whilst retaining reasonable green spaces and native vegetation cover (Batáry et al. 2018) and increasing forest edge effect through within-habitat heterogeneity (Marzluff 2005, Alberti 2015, Padayachee et al. 2017).

Nonetheless, loss of native habitat land-cover (e.g. forest) is strongly associated with local extinctions (Marzluff 2005); coming with a dispossession of a variety of resources and breeding sites (García-Arroyo et al. 2020). Anthropogenic settlements and commodities (e.g. transportation channels; roads) and agricultural/horticultural lands cleave and perforate natural habitats; creating a patchy and fragmented habitat (Marzluff 2005; Groffman et al. 2014; Fraser et al. 2019). This severing of connections complicates dispersal, particularly penalizing less mobile organisms (Alberti 2015). Habitat loss and fragmentation also poses a problem for high trophic level predators, which typically require large territories, with sustained prey populations. Moreover, humans have tampered extensively with the trophic ecology of human-inhabited regions, removing predators or sequestering land from herbivores for pastures (Nilsen et al. 2007, Jiang et al. 2016, Fraser et al. 2019). Urbanization also comes with new threats (e.g. window collisions, cat predation, moving cars) (Santori et al. 2018, Mella-Méndez et al. 2019, García-Arroyo et al. 2020) and. The resulting depauperate assemblages are dominated by few dominant urban exploiters, usually synanthropic natives or NNAS (Shochat et al. 2010, García-Arroyo et al. 2020).

As a result of urban selective pressures and the expansion of invasive NIS, biotic homogenization is occurring globally (Kark et al. 2007, Newbold et al. 2018). Congruent with large-scale habitat homogenization, this results in increased opportunity of conditions fitting for biological invasions. When the biota of different urban ecosystems become nearly interchangeable, this complicates the conservation of native species (Simberloff and Holle 1999, Collins et al. 2000, Groffman et al. 2014, Wang et al. 2021).

Biotic interactions: established NNAS and exotic vegetation

Biotic interactions between NNAS are an important aspect of the colonization process, as they may facilitate invasions predominantly through indirect effects, but also via direct trophic (mutualistic, commensal) effects (Simberloff and Holle 1999, Alberti 2015, Cadotte et al. 2017): a process termed invasional meltdown (Simberloff and Holle 1999). Invasive NNAS act towards impoverished and simplified communities of low biodiversity, by displacing and/or removing natives and altering ecosystems and their functions which, simultaneously with the direct effects of urbanization on biodiversity (Walters 2006, Short and Petren 2012, Mbenoun Masse et al. 2017, García-Arroyo et al. 2020), may create conditions exploitable by other NNAS (Cadotte et al. 2017). Direct competitive interactions of established NNAS with natives are widely reported (Simberloff and Holle 1999, Tennessen et al. 2016, Taggar et al. 2021). For instance, population density of invasive house sparrow was negatively associated with native avian species richness across four urban settlements in western Mexico. House sparrows compete for nesting opportunities and food resources by superior aggressiveness, and can be sources of pathogens such as avian pox and malaria (García-Arroyo et al. 2020). The tadpoles of American bullfrogs can outcompete native tadpoles, forcing skewed realized niche and alterations of their microhabitat that render them more susceptible to predators, whilst the adults are generalist that can predate also on other native amphibians (Ficetola et al. 2007).

Habitat modification by NNAS that are ecosystem-engineers may directly render the habitat unfavourable for natives, or ignite a cascade of changes leading to a similar outcome and consequently creating opportunity for invasive species (Simberloff and Holle 1999). Asian water buffalo introduced to northern Australia in the 1800s (Skeat et al. 1996) as life-stock and beasts of burden, quickly became feral and dispersed. Altering habitat hydrology and plant communities of forests and flood plains, and compacting soil, promoted domination by an already invasive Central American shrubby legume Mimosa pigra, ultimately damaging local fauna dependent on the native sedgelands habitat. Another example is the zebra mussel Dreissena polymorpha, invasive to the Great Lakes region of the USA since the 1980s. Being a highly efficient filter-feeder, Dreissena cause a decline in phytoplankton biomass whilst achieving deposition of rich excreted organic material. The new soft and deposit-feeder-friendly benthic substrate is improper to some native inhabiting invertebrates, but beneficial to others, such as the invasive Eurasian faucet snail Bithynia tentaculata (Simberloff and Holle 1999).

NNAS profit from exotic vegetation for familiar food and shelter (Gering and Blair 1999, Wilke et al. 2018). For instance, invasive fructivores in ranges lacking native fleshy fruit-bearing vegetation may feed from exotic horticultural plants (Collins et al. 2000, Cesari et al. 2018). Florida invasive red-whiskered bulbuls utilize the same nesting and feeding resources that they would in their native range of tropical Asia, thanks to the extensive anthropogenic modification of urban flora (Simberloff and Holle 1999). The popularity of highly attractive exotic ornamental plants attracts exotic pollinator species with shared eco-evolutionary history, whilst afflicting native pollinator populations. Although the plant species tend to be most alluring to both native and invasive pollinator species, accessibility of flowers and quality of the nectar or pollen are often incompatible with native pollinators, inducing them to only hover around but not enter and pollinate flowers. This negative impact on pollination network is particularly meaningful in highly urbanized settings, where density of both plant and pollinator species is already low (Buchholz and Kowarik 2019).

Biotic resistance and relaxation of predation pressure

Overcoming biotic resistance (the suppression or prevention of NNAS by predation and competition from natives) is key for successful establishment of NNAS (Carthey and Banks 2018, Santana Marques et al. 2020). Though NNAS may no longer need to deal with their native-range predators, they will have to respond to predation pressure from unfamiliar local predators. NNAS will not have shared evolutionary history with native predators making them vulnerable due to naiveté and lack of anti-predator behaviour (Carthey and Banks 2018, Twining et al. 2020). Biotic resistance controls the spread of NNAS in natural areas, but in cities the diversity and abundance of native competitors and predators is usually low (Hernández-Brito et al. 2020, Santana Marques et al. 2020). Despite the presence of invasive non-native predators (e.g. cats) (Sims et al. 2008, Shochat et al. 2010, Carthey and Banks 2018), predation pressure generally relaxes with increasing urbanization (Friesen et al. 2013). Studies such as Gering and Blair (1999) or Roshnath et al. (2019) indicate that predation pressure on avian nests is lower in urban environments than in semi-natural or natural areas. This can explain the high abundance of prey avian invasive NNAS (e.g. European starling, house sparrow, feral pigeon) in cities (Gering and Blair 1999). Urban birds and mammals (e.g. squirrels) also forage more freely and efficiently, rather than only foraging close to shelter suggesting perceived safety from predation (Shochat et al. 2010). The urban landscape can also be heterogeneous, there is often variability in predation pressure within the same city (Gering and Blair 1999), which might offer refuge from predators (Hernández-Brito et al. 2020).

Predator release profusely aids biological invasions (Hernández-Brito et al. 2020). In guppies (Poecilia reticulata), absence of predators has resulted in larger body size, enabling specimen to travel longer distances and thus facilitating dispersal. Larger body size also allows better persistence in difficult urban streams, and for females it is associated with higher fecundity and more offspring (Santana Marques et al. 2020). Predation relaxation allows NNAS to develop facilitative interactions with native species, granting safety from predation also in areas with biotic resistance. Invasive alien monk parakeet in central Spain have begun nesting close to active nests of large native white stork, reducing predation risk from raptors, and enabling spread into rural areas (Hernández-Brito et al. 2020).

Despite the evidence we presented, the extent of the relevance of biotic resistance in urban settings is debated. In some systems the native predators mostly ignore invaders, while in other scenarios predation by natives has little influence on invader success (Hernández-Brito et al. 2020, Santana Marques et al. 2020). For example, the presence of other fish species in urban streams showed negative but insignificant effects on guppy traits and density. This is explained by compensatory life-history traits and high food availability (Santana Marques et al. 2020). Moreover, in some NNAS a high marginal value of food in urban landscapes may induce bold behavioural syndrome, where anti-predator responses become weakened or absent even if the predator is recognized. For example, Carthey and Banks (2018) found invasive peri-urban black rats to be naïve of unfamiliar native predators but also unresponsive to recognized exotic predators, when rushing for valuable urban food (ie peanuts) (Carthey and Banks 2018). Furthermore, as noted earlier, habitat heterogeneity and variability in predation pressure amongst sites (Gering and Blair 1999) can create refuge from predation (Hernández-Brito et al. 2020).

Conclusions and future outlooks

Biotic and abiotic characteristics of the urban landscape facilitate biological invasions through an interacting and complex set of drivers. Starting with increased NNAS propagule pressure by high rates of intentional and unintentional introductions (Hulme et al. 2008, Banha et al. 2017), trait-selected non-indigenous animal species take advantage of increased resources, interspecies mismatches and often absent or weak biotic resistance (Kark et al. 2007, Grarock et al. 2014, García-Arroyo et al. 2020, Hernández-Brito et al. 2020, Santana Marques et al. 2020) due to the removal of native predators and competitors in urban areas (Groffman et al. 2014). Furthermore, the milder urban climate, the presence of artificial structures and exotic flora may facilitate the establishment and spread of NNAS (Simberloff and Holle 1999, Collins et al. 2000, Groffman et al. 2014, Cordonnier et al. 2020, Barsotti et al. 2021). This is further aided by interspecific interactions between NNAS and between habitat and established ecosystem-engineer NNAS (Simberloff and Holle 1999). Meanwhile, urban expansion and global homogenization broaden invasive potential of NNAS to a wider range of locations (McKinney 2006, Groffman et al. 2014, Barsotti et al. 2021). Inter-cities connections aid dispersal of established NNAS populations (Botham et al. 2009, D’Amico et al. 2013, Sherpa et al. 2020). Simultaneously, urban refuge conditions allow for development of adaptations that may also spread to the natural non-native environment (Candler and Bernal 2015, Hernández-Brito et al. 2020, Baxter-Gilbert et al. 2021).

With rising urban expansion and globalization, biological range shifts and their destructive natural, economic and health impacts are a growing concern (Knudsen et al. 1996, Kovacs et al. 2014, Parejo et al. 2015, Padayachee et al. 2017, Mori et al. 2018a, Hima et al. 2019, Gethöffe and Siebert 2020, Godefroid et al., 2020, Diagne et al. 2021). Understanding of NNAS and their interactions with the abiotic and biotic environment is crucial to developing strategies to minimize the arrival, establishment and spread of NNAS, and thereby of their ecological impacts (Sato et al. 2010, Cadotte et al. 2017, Padayachee et al. 2017). Still, well-designed studies are lacking, as well as studies investigating socio-economic spatial pattens, or considering the sub-urban environment as a separate and different class from the urban environment (Cadotte et al. 2017). Moreover, our literature search finds the Global South (i.e. Africa, South and South-Eastern Asia, South America and the Middle East) to be overall remarkably underrepresented. Another important knowledge gaps requiring research attention would be to understand how adaptive behaviours evolved from exposure to humans and urbanization, such as boldness (Candler and Bernal 2015, Baxter-Gilbert et al. 2021), may be translating into the natural environments surrounding invaded urban landscapes, and modify interspecific dynamics with native biota, potentially increasing invasive potential to previously resistant natural landscapes.

Aknowledgements

We thank two anonymous referees for insightful comments on a previous version of this manuscript, Dr. Caroline Isaksson for feedback and generous support on the first draft produced, and Dr. Federico de Pascalis (ISPRA) for prescious guidance towards the methods.

Author contributions

Eugenio Carlon (Conceptualization [lead], Investigation [lead], Methodology [lead], Project administration [lead], Visualization [lead], Writing—original draft [lead], Writing—review & editing [lead]) and Davide M. Dominoni (Supervision [supporting], Writing—review & editing [supporting])

Supplementary data

Supplementary data is available at JUECOL online.

Conflict of interest: None declared.

Funding

Davide M. Dominoni was funded by a NERC Strategic Call grant (NE/W005042/1).

References