Factors Shaping Biodiversity in Urban Voids: A Systematic Literature Review

Abstract

Urban voids, such as vacant lots, brownfields, fallow land, wasteland, and the spaces between buildings, have the potential to serve as habitats for a diverse range of plants. However, their value for plant conservation remains understudied because of their informal and neglected nature. The aim of this systematic review is to consider the potential for urban voids to contribute to urban plant biodiversity. A total of 55 studies from 14 countries were analysed for geographical trends, factors influencing plant biodiversity, research methods, and relevant environmental parameters. The results of this study show that the factors influencing the plant diversity of urban voids can be divided into three main categories: biophysical, temporal, and landscape factors. Biophysical factors, such as size, age, soil, and vegetation structure, as well as temporal factors, including site management and human interventions, are the most important at the site-level scale. In contrast, landscape factors, such as the location and distance from other green spaces, are more dominant at the larger urban scale. The review has identified a significant limitation in the available literature, with relatively few studies examining the influence of urban voids’ shape, ownership, and microclimate on biodiversity. With respect to research methods, the results suggest a recent increase in the use of GIS-based and remote sensing techniques to investigate the plant diversity of urban voids. The findings discussed in this paper indicate the need for further research to comprehensively understand the factors that promote urban vegetation diversity in urban voids.

1. Introduction

Urbanisation and its associated land-use modifications and environmental disturbances have a profound impact on natural habitats and the communities that inhabit them [1,2]. These impacts include an elevated pressure on local species diversity, leading to potential declines [3], as well as the formation of novel ecological communities [4]. While urban areas have been recognised as sites with high spatial heterogeneity in terms of green spaces, supporting a substantial number of plant and animal species, they also play a vital role in biodiversity conservation [5]. However, the detrimental effects of urbanisation on a larger scale are evident in the phenomenon of functional homogenisation, characterised by a reduction in beta diversity. This phenomenon is characterised by the decrease in specialist local plant species, which are frequently substituted by generalist species that exhibit greater adaptability to diverse urban environmental conditions [6,7]. Nevertheless, it is important to acknowledge that the introduction of numerous exotic species within cities, along with the decline of native species, frequently gives rise to biotic differentiation, especially in certain urban spaces that experience lower levels of disturbance or have a more natural character, where the presence of this mix of exotic and native species can lead to distinct ecological communities and increased biodiversity [8].

Urban plant biodiversity refers to the variation in plant species within urban ecosystems, which play a crucial role in providing essential services that benefit urban residents [9]. These ecosystems, including those that arise after human disturbances such as abandoned residential lots or industrial brownfields, commonly referred to as urban voids in this study, offer recreational spaces and contribute to the preservation of biodiversity [10,11,12]. It is worth noting that while vegetated urban voids, characterised by unused lands covered with spontaneous vegetation and minimal human intervention, are often dominated by non-indigenous species, several studies have demonstrated their capacity to support a substantial richness of native species and also to provide significant habitats for rare species [13,14]. Urban voids can exhibit an even greater diversity of native species compared to semi-natural urban habitats, such as emerging urban forests [15]. Therefore, considering this potential and their widespread distribution in certain urban landscapes, some studies suggest that these urban voids can serve as viable habitats that complement urban biodiversity conservation networks [13,16].

The term “urban void” (UVs) has evolved within various disciplines in recent decades, originating in the United States in the 1970s and subsequently gaining prominence in the United Kingdom and European cities during the post-industrial era of the 1990s [17,18]. Prior to the 70s, key figures and events stand out for their significance to the study of ecology in city contexts. In the United Kingdom, works such as Salisbury’s “The Flora of Bombed Areas” drew attention to the succession processes in the toxic, devastated voids in wartime London and subsequently throughout Europe when voids would become an inherent characteristic of the postwar landscape. However, in general, work on urban voids, and biodiversity has been sporadic, according to figures from the United Kingdom, such as O.L. Gilbert [19], who argued that urban ecosystems represent a critical frontier for ecological research where human influence is not merely a factor but a dominant force. Had ecology’s pioneers focused on cities rather than pristine environments, anthropogenic impacts would have been central to foundational ecological theory. Gilbert notes that throughout the twentieth century, a “shortage of detailed ecological studies carried out in highly urbanised areas means that only incomplete evidence and scattered examples are available to illustrate anthropogenic influences at work”.

In Germany, Sukopp greatly advanced the understanding of biodiversity, landscape factors and urban voids or “Stadtbrachen” in German. Sukopp, a pioneering German ecologist, is widely recognised for making fundamental contributions to our understanding of urban ecosystems and their ecological dynamics, particularly in conceptualising urban voids as critical sites of biological diversity. His groundbreaking work during the 1970s and 1980s challenged prevailing perceptions of cities as biological deserts, instead proposing that urban landscapes—especially neglected interstitial spaces—represent complex ecological systems with significant biodiversity potential [20,21,22].

While Sukopp’s research in Germany demonstrated that marginal urban spaces like vacant lots, industrial brownfields, and transitional areas between built environments could serve as important habitats for plant species, contemporaneous researchers in France were exploring similar ecological themes. The long career of Paul Jovet and colleagues such as Bernadette Lizet [23,24,25,26] expanded the contexts and study of ecology to the city, drawing attention to late twentieth-century post-industrial sites such as the Bercy Warehouses [27], which were redeveloped as parklands and also for contemporary mixed uses. Other early works by Coquillat [28] are also of note for foregrounding prosaic urban contexts, such as pavements, as sites of ecological succession.

The work of Jovet, Sukopp, and others laid the theoretical and methodological groundwork for contemporary urban ecological research, inspiring generations of landscape architects, urban planners, and conservation biologists to recognise the ecological value of urban voids and integrate biodiversity considerations into urban design and management strategies.

For instance, European landscape architects have since investigated the design possibilities of spontaneous vegetation in the urban fabric [29,30]. In this way of thinking, Gilles Clément, one of France’s most well-known contemporary landscape architects, advanced the concept of the “Third Landscape” [31] and provided a theoretical framework concerned with appreciating neglected urban spaces as biodiversity havens. Clément’s “Jardin en Mouvement” [32], situated in Andre Citroen Park and created in the late 1990s, contrasts with conventional gardens where plants occupy fixed positions. Instead, the experimental plot functions as a “vacant lot” that combines both garden species and spontaneous flora typically found in Parisian terrain vague spaces. The garden’s management philosophy deliberately incorporates natural vegetative succession processes and allows herbaceous plants to naturally appear and disappear, creating an ever-changing spatial dynamic that suggests new pathways for visitors to explore. The “moving garden” represents a synthesis of urban ecological principles and insights drawn from the study of vacant lots, offering a new paradigm for understanding the potential of urban voids as dynamic, evolving spaces rather than static environments. In Germany, Peter Latz is perhaps the most notable example, with his landmark work on Landschaftspark Duisburg-Nord [33,34,35,36]. This former industrial site was transformed into a public park that deliberately incorporated existing industrial structures and spontaneous vegetation. Rather than clearing and replanting the site, Latz used ecological processes to take advantage of the recruitment of spontaneous vegetation.

Other international researchers, such as Ignatieva, have extended Sukopp’s pioneering work on urban voids by globalising the perspective of urban ecological research. Where Sukopp primarily focused on German urban landscapes, Ignatieva has examined urban biodiversity across continents and cultures, comparing ecological dynamics in Russia, Scandinavia, Australia, and New Zealand, for instance. Ignatieva’s concept of “third nature” complements Sukopp’s earlier observations about the ecological potential of neglected urban spaces [37,38].

Her work has examined, for instance, the complex interactions between spontaneous vegetation, urban design, and ecological systems. Ignatieva’s contributions have been particularly insightful in revealing how unplanned spaces contribute to biodiversity conservation in Australia and New Zealand [37,38], furthering our understanding of urban voids as dynamic ecological niches.

Urban voids (UVs) have emerged as a focus of Chinese research more recently. In the past few decades, ecologists such as Changdu Chen [39,40,41] developed research on urban plant communities and biodiversity following time collaborating with Sukopp, which resulted in studies on Tianjin’s vegetation patterns [42]. Chen’s conception of landscape as the “fourth dimension” of biodiversity conservation has guided his extensive plant surveys throughout China [43,44,45,46].

In China, research on UVs gained theoretical depth through Shijun Ma and Rusong Wang’s “social-economic-natural complex ecosystem” framework, which articulated the role of urban voids within larger-scale urban ecosystems [47]. Wang, in particular, emphasised structural diversity as key to system stability, positioning biological resources in urban spaces as an “ecological technology” for regeneration [48,49]. Planning researchers such as Youren Wu [50] demonstrated that sound land use policies are essential for environmental protection amid China’s rapid urbanisation. Complementing such studies, Liangyong Wu’s research advocated for better planning of undeveloped urban land to enhance efficiency and environmental quality [51]. While not explicitly referencing the concept of “urban voids”, such planning scholarship helped lay the foundations for subsequent research on the subject.

Within China, disciplinary studies in landscape ecology provided the theoretical bridge linking UVs to biodiversity by illuminating how spatial structure influences ecological function. Scholars, including Bojie Fu, Jianguo Wu, and Jinping Guo [52,53,54], explored such relationships in their work, and by the late 1990s, Duning Xiao’s research on suburban landscape patterns demonstrated the ecological significance of “urban fringe” areas. Such transitional zones, exhibiting high biodiversity due to edge effects, were recognised as valuable components of the urban ecosystem [55,56,57].

China’s most well-known landscape architect, Kongjian Yu, has worked to link science and theory to design projects through the concept of “landscape ecological security patterns.” Site-specific works, such as the landmark Zhongshan Shipyard Park, were reminiscent of Latz’s German approach. In this project, Kongjian Yu transformed the site through new interventions while integrating existing post-industrial structures along with remnant and spontaneous vegetation [58,59,60].

Urban voids are often perceived negatively and are typically associated with higher crime rates and decreased property values [61]. However, as the above-outlined research demonstrates, they can also be viewed as valuable resources for local environments and represent a new type of urban space that addresses the needs of diverse urban populations and enhances urban ecology [16,62]. With ongoing urbanisation, the number of urban voids is increasing, necessitating better understanding and management of these urban spaces. Rather than seeing such spaces as a background to urbanisation and of secondary ecological importance, recent research by Gandy [63] and others [15,64,65] emphasises their importance, noting that so-called “wastelands” make fundamental contributions to urban ecology [66].

Despite their diversity, both within cities and around the world, UVs have generally been used as an abstract neutral term to describe a number of types of landscapes, including vacant land, terrain vague, urban wilderness, informal green space, wastelands and leftover space, which often causes inconsistency and confusion [67,68]. The lack of a clear definition and standardised terminology for UVs in research hinders progress in this field and may result in overlooking relevant prior studies exploring these urban spaces. There have been some attempts to systematically review and reconcile this varied research, including the helpful annotated bibliography by Lopez-Pineiro [69]. Without clear definitions or standardised characterisation methods, it is challenging to compare research findings, build knowledge, and establish consistent global parameters. To address this issue and provide a sense of coherence on this topic, this study uses a preliminary, non-exclusive definition of ‘urban void’ (UV) developed from a reading of Lopez-Pineiro, which can be described as follows:

“a landscape element (patch or corridor), which consists of any unutilized, uncultivated, undesigned, non-built up, or abandoned land and premises which exist in urban areas, that at least partly covered with non-remnant, spontaneous vegetation showing signs of lack of maintenance, or presenting bare soil. The term should not be confused with open spaces that serve specific purposes, like parks or sidewalks.”

A few review articles have examined the role of UVs in urban biodiversity conservation [15,70,71]. While these reviews have provided valuable insights into the factors influencing UV biodiversity and strategies for incorporating UVs into urban planning, there is a need to explore emerging patterns that have not been extensively covered. As such, an understanding of the character of urban voids has been overlooked, with leading figures such as Nassauer suggesting the area remains a research frontier [72].

This review aims to fill this gap by examining recent research in UV biodiversity, specifically focusing on urban voids’ plant community diversity. By also including studies conducted in Chinese, the findings will have greater relevance and applicability to urban spaces in developing countries. Additionally, this review provides a comprehensive categorisation and summary of the factors influencing UV biodiversity, addressing methodological approaches and study durations, which significantly impact the analysis and conclusions drawn from UV biodiversity data.

To achieve these objectives, this review addresses the following key questions: (a) What are the patterns and trends of current UV plant diversity studies across the globe? (b) What factors or parameters of UVs are most influential when it comes to plant biodiversity? (c) What factors or parameters of UVs are more influential on the landscape connectivity of plant communities? (d) What are the different methods and software employed to quantify the impact of these factors on plant diversity in UVs? By answering these questions, this review aims to identify knowledge gaps and suggest future research directions. Through a systematic review of peer-reviewed scholarly studies on the plant diversity of UVs, this paper intends to assist researchers and practitioners in evidence-based management of UVs to conserve or restore urban plant diversity. The study includes no limitations on geographic locations, aiming to provide guidance for planning and management strategies regarding UVs and have important policy implications for the conservation of plant diversity in urban areas.

2. Materials and Methods

This study applied a systematic review approach [73,74] enhanced by a targeted critical review. A systematic review typically focuses on geographic, theoretical, and methodological aspects by analysing the current trends in the literature, which is helpful in identifying current knowledge gaps and establishing more appropriate frameworks for future research. Based on the author’s proficiency in multiple languages, this review included Chinese and English papers to extend the scope of the review.

A systematic search was conducted using three major databases, namely Science Direct, Google Scholar, and China National Knowledge Infrastructure (CNKI). This study combined the search terms within the categories of ‘UVs Keywords’ and ‘Plant Keywords’ using the Boolean operator ‘OR’ (

Table 1. Search terms used in English and their translation in Chinese.

UVs Keywords	English	Chinese
	Urban voids	城市空墟地
	Urban vacant land	城市空地
	Informal green space	非正式绿地
	Urban brownfield	城市棕地
	Urban wilderness	城市荒野
	Wastelands	闲置地/废弃地
Plant Keywords
	Plant diversity	植物多样性
	Plant species richness	植物物种丰富度

). The two categories were then combined using the Boolean operator ‘AND’, for example, [“Urban voids” OR “Urban vacant land” OR “Informal green space” OR “Urban brownfields” OR “Urban wilderness”] AND “Plant diversity”. The systematic searches were conducted in late 2022, and no time limit was applied to the search. While the systematic search was conducted in 2022, subsequent critical target searches were completed in 2025 to contextualise the research historically in terms of early papers and very recent papers.

A total number of 965 publications were found during the initial search process, which went through an additional three-step evaluation process for final inclusion, and the total number of records after exclusion was 55 (Supplementary Material S1; Figure 1). Firstly, they were evaluated based on the title, followed by an assessment based on an abstract for potentially relevant articles. Finally, the articles that were still potentially relevant based on their abstracts were evaluated by their full text for inclusion based on the inclusion criteria outlined below. For publications to be considered for inclusion in this review, they were required to meet each of the following criteria:

• Relevant subjects: The case study site should be recognised as UVs as defined in this review.
• Relevant type of factors: Studies should have identified, described, or quantified specific UV-related parameters such as site attributes, human intervention level, or urban landscape context.
• Relevant type of data: Sufficient data on UVs and plant diversity parameters were provided by the study for literature trends analysis.
• Relevant type of outcomes: Measures on the diversity of plants should have been provided.

In cases where uncertainty arose during any stage of the screening process, the reviewer leaned towards including the article.

This study systematically analysed the literature to identify the role of UVs in increasing urban plant diversity by using a taxonomy of thematic aspects summarised in

Table 2. Categories and their definitions used to group the various factors of UVs.

Category	Definition
Biophysical	The environmental components of vacant urban sites, encompassing both non-living elements and living components that shape ecosystem function. These site-specific characteristics form the ecological foundation upon which plant communities establish and develop [75].
Temporal	Time-related dimensions and management interventions that evolve throughout an urban void’s existence. These capture the dynamic nature of urban voids and their transformation through both planned and spontaneous succession processes [33].
Landscape	The contextual and inter-site relationships of fragmented urban open spaces within the metropolitan matrix. These spatial determinants describe how isolated patches interact with surrounding urban fabric and contribute to broader ecological networks [76].
Scale	Refers to the analytic spatial framework at which urban void biodiversity is studied and analysed. These factors address the different levels of resolution at which researchers examine ecological patterns and processes within urban voids [77].
Method	Encompasses the specific research approaches and techniques used to study urban void biodiversity. These factors describe the procedural aspects of how data on urban voids is collected, processed, and analysed [78].
Software	Refers to the computational tools and programs used for data analysis, spatial mapping, and statistical modelling in urban void biodiversity research. These tools help researchers process complex datasets and visualise spatial relationships [79].

and

Table 3. Taxonomy of UVs with definitions of the various factors discussed in the literature.

Category	Sub-Category	Factor	Working Definition of Factors
Environmental	Biophysical	Size/Shape	Physical dimensions and geometric configuration of urban voids that influence habitat area and edge effects [9].
	Biophysical	Age	Current ecological successional stages of a site since abandonment and transition from previous uses [80].
	Biophysical	Soil	Chemical and physical properties of soil including nutrient content, pH, metal concentration, and depth [81,82].
	Biophysical	Land Use Legacy	Soil, site and vegetation conditions resulting from past uses [83,84].
	Biophysical	Microclimate	Local atmospheric conditions including temperature, humidity, and light exposure [5,85].
	Biophysical	Plant Species	Existing vegetation composition that influences further plant colonisation and community development [86].
	Temporal	Accessibility	Frequency to which humans enter and interact with urban voids in everyday routines [72].
	Temporal	Human Disturbance	Level and frequency of human interventions including trampling, waste dumping, and management activities [12].
	Temporal	Restoration Strategy	Deliberate approaches to enhance biodiversity, such as seeding, planting, or habitat creation [87].
	Landscape	Connectivity	Degree to which urban voids are physically or functionally linked to other green spaces [13,16].
	Landscape	Context/Location	Position within the urban–rural gradient and surrounding land use types [88].
	Landscape	Distance to Green spaces	Proximity to other vegetated areas that can serve as seed sources [89].
Methodological	Scale	Macro	City-wide or regional level analysis of urban void networks [6].
	Scale	Local	Neighbourhood or district level assessment of urban void clusters [90].
	Scale	Micro	Site-specific evaluation of individual urban voids [6].
	Method	Study Period	Duration and timing of research observations [91].
	Method	Case Study	Analysis of specific urban void sites and their characteristics [92].
	Method	Experimental	Controlled manipulation of variables to determine causal relationships [81].
	Method	Remote Sensing	Use of satellite or aerial imagery to assess urban void distribution and properties [93].
	Method	Literature Review	Synthesis of existing research on urban void biodiversity [15].
	Method	Statistical Analysis	Quantitative approaches to analyse relationships between factors and biodiversity [94].
	Software	R (version 2.11.1-4.1.2)	Programming language used for statistical analysis and modelling [95].
	Software	ArcGIS (version 10.2-10.6.1; QGIS (version 2.12-3.0)	Geographic Information Systems used for spatial analysis and mapping [96].
	Software	Others (e.g., Google Earth Pro 7.1; i-Tree 5.0; SPSS 16.0–23.0; MATLAB R2017a)	Additional software tools employed for specific analytical purposes [81].

. Environmental categories, including biophysical, temporal, and landscape dimensions, were considered, as were methodological categories, including scale, method and software. These broad thematic categories, set out in Table 2, were further divided into factors, and these are set out and defined in Table 3. These themes and factors formed the main content of the analyses conducted.

Qualitative, quantitative, and graphical analyses were performed using the above factors to present and synthesise findings from the finally selected 55 papers and were systematically analysed to identify geographical, theoretical and methodological trends and gaps in the scholarly literature, adopting the Rupprecht et al. [71] approach, which included multi-language resources, systematic literature search via search logic table and Boolean functions and multi-scale screening of paper contents. One of the limitations of Rupprecht was its lack of Mandarin/Chinese references, and this paper addresses this limitation through its inclusion of Chinese literature. Analyses were completed for the following themes, and these are set out and presented in the results section:

• Geographical distribution of studies
• Analysis in relation to factors influencing urban void plant diversity

  ○

  Physical factors discussed in reviewed studies

  ○

  Temporal factors discussed in reviewed studies

  ○

  Biophysical factors discussed in reviewed studies

  ○

  Methodological factors discussed in reviewed studies

3. Results

3.1. Identification of Relevant Studies

Based on the systematic methods outlined in the previous section, we identified 965 articles from the three major databases. Of these articles, a total of 55 publications (see Supplementary Material S1) were selected for final review after screening across 27 journals (See Figure 2). The selected publications exhibited a diverse distribution among various journals, with Landscape and Urban Planning and Urban Forestry & Urban Greening being the most prominent, each accounting for 13% of the total papers (7 papers). Urban Ecosystems contributed five papers (9%), while Sustainability, Applied Vegetation Science, Biological Conservation, and Landscape Architecture Frontiers (景观设计学) each contributed three papers. Additionally, the Journal of Applied Ecology and Chinese Landscape Architecture (中国园林) each contributed two papers, while the remaining journals yielded one paper each. Most of the selected papers were written in English, with only six in Chinese. The years of the publications ranged from 2005 to 2022.

3.2. Geographical Distribution of Studies

The review included publications from 14 countries, mostly published in peer-reviewed journal papers. The United States made the largest contribution (27%) with 15 papers, followed by China, which contributed 25% with 14 papers. Germany contributed 18% with 10 papers. France contributed 7% with four papers, and the United Kingdom contributed 5% with three papers. The remaining highlighted countries each had one publication, and mostly were located in Europe. These findings align with Bonthoux et al.’s [15] observation that there is a geographical concentration of research in the United States and Europe. However, it is worth noting that the increasing number of papers from China may be attributed to the inclusion of research papers written in Chinese during the literature search. The geographical context analysis in this study includes the case study locations and institutional research interests. The overall geographic distributions of the selected papers are depicted in Figure 3, highlighting the predominance of publications from the United States.

3.3. Factors Influencing Urban Void Plant Diversity

As outlined in the methods section, reviewed studies were classified according to the factors they focused on, and this classification is visualised in Figure 4. The majority of the studies included in this review (N = 40, 73%) examined the effects of biophysical factors, while 31 studies (56%) reported on the effects of landscape factors on UVs’ plant diversity. A total of 28 studies (51%) examined the effects of temporal factors. Many of the reviewed articles did not exclusively focus on one type of factor. Approximately one-third of the articles collectively addressed biophysical and landscape factors, while a similar proportion jointly discussed biophysical and temporal factors. However, only a limited number of articles explicitly mentioned temporal and landscape factors together (N = 2, 4%) or incorporated all three categories of factors (N = 4, 7%) in their analysis.

3.3.1. Biophysical Factors Discussed in Reviewed Studies

As defined in the methods, biophysical factors of UVs refer to both the living and non-living components of the site that shape the ecological community and its development. These factors and their importance to plant diversity have been most frequently discussed by scholars (N = 40, 73%). This review identified seven groups of physical factors linked to the UV plant diversity among these studies: size (N = 10, 18%), age (N = 8, 15%), shape (N = 3, 5%), soil condition (N = 22, 40%), type of surface (N = 4, 7%), microclimate conditions (N = 3, 5%), and finally the land-use legacy (N = 10, 18%). Figure 5 provides a summary of the number of studies exploring the impact of physical factors on plant species richness and taxonomic composition in UVs (see Figure 5).

In selected studies, it was found that the plant species richness in UVs is positively correlated with the size of the area [9,13]. However, Vega and Kuffer [97] argued that smaller landscape patches had higher beta diversity than larger patches. Additionally, UVs with linear shapes, such as railway edges, were found to have higher plant beta diversity, which means they have higher spatial variation in species composition among communities at regional or landscape scales as they allow seed spreading in a radial pattern [98]. Other factors that have been found to be positively related to plant species richness in UVs include the UV’s age, which is closely related to the ecological succession of the site [9,13,99], diversity of soil nutrient content [82], soil metal concentration [80,100], and soil depth [101].

On the other hand, the presence of hard surfaces and substrates, such as concrete and asphalt, has been found to be negatively related to UV plant species diversity [82,85,102]. Land-use history and the legacy of UVs can also cause variations in the plant community composition as well as species richness [72,103,104]. Johnson et al. [83,84] supported this argument by comparing the plant species diversity in remnant gardens and former building footprints in vacant lots. Microclimate variables, such as air temperature and humidity, could influence the presence of certain species in UVs, as discussed by Godefroid, Monbaliu and Koedam [82] and Blouin, Pellerin and Poulin [5]. There are still some biophysical factors that have not been included in the scope of discussion in the selected articles, such as ownership and bird and insect communities on site, which may also influence plant diversity in UVs. Future research is required to investigate these biophysical aspects, which are lacking in the current research.

More than half of the selected papers (N = 34, 62%) examined the value of UVs in comparison to other urban green spaces. These studies consistently reported a high number of plant species across different types of UVs, with some UVs even harbouring rare plant species [80,97,105,106]. Moreover, investigations have extended to comparing UVs with habitats located in rural areas to comprehensively evaluate the contribution of UVs to plant diversity. The findings indicate that UVs exhibit greater plant species diversity than vacant areas in rural settings [107,108]. However, the disparity may be attributed to the presence of non-indigenous species [109].

In regard to the types of plants, the focus of UVs’ plant diversity was primarily on general herb and shrub species (N = 33, 60%). Other researchers also studied specific subsets of vegetation, such as vascular plants (N = 2, 4%), as well as groups of species that did not belong to a specific taxon, such as spontaneous or non-native vegetation (N = 5, 9%). In terms of other plant species, woody plants, or young forest (N = 4, 7%) and grassland (N = 1, 2%) were also studied.

3.3.2. Temporal Factors Discussed in Reviewed Studies

As previously defined, temporal factors refer to aspects of UVs that occur over time and have a duration. Based on the selected studies, temporal factors include the following sub-categories: Human disturbance (N = 19, 35%), restoration strategies (N = 12, 22%) and accessibility (N = 1, 2%) (Figure 6). Human disturbance refers to regular or irregular management activities, as well as other human-induced activities such as trampling and dumping [95,110]. Restoration strategies encompass various approaches, such as seeding and proposed restoration frameworks, which have the potential to impact the plant diversity of UVs. Lastly, accessibility pertains to the measure of a UV’s frequency and ease of access to the public, wherein factors like fences can restrict the ease of reaching UVs. Various reviews also identified studies that have assessed the influence of visiting and management frequency on UV biodiversity [15,64,111]. However, the reviewed papers either discussed the impacts of human presence or examined the effects of site management on UV plants without systematically differentiating intentional management from unintentional human disturbance. In addition, some planting strategies and planning principles that may affect UVs’ plant diversity have not been mentioned or discussed in these articles. This literature review distinguishes between different types of human intervention to UVs and considers the restoration strategies proposed in previous studies as one of the temporal factors in order to provide a more comprehensive understanding of the temporal factors of UVs.

Human disturbance, such as vegetation removal, is the most common site management practice. Some studies have shown that appropriately reducing the frequency of management can increase the number of plant species [65,94,112], especially for herbaceous plants [113,114]. However, there are conflicting results in the literature, with some studies finding that medium intensively managed plots have the highest plant diversity [110], and few other studies found no discernible effect between mowing and plant diversity in UVs [115].

Similarly, limited access to UVs may also have a negative impact on plant diversity of UVs [72,87], and fences built in UVs may discourage some human disturbances, such as transport, trampling, and dumping. Soil transport caused by traffic in and around UVs may increase the diversity of both alien and endangered plants [109], while the trampling intensity may be negatively related to the site plant diversity [101]. Trash is another common human disturbance in UVs, and some studies have revealed intriguing findings regarding UVs with higher amounts of trash. It was observed that vacant lots with increased trash exhibited a smaller vegetation area yet displayed higher species richness and evenness [87]. This phenomenon can be attributed to the fact that UVs with higher levels of trash tend to receive a greater influx of wind-borne seeds. Consequently, this influx contributes to an increased richness of plant species within such UVs. [72].

In addition to the human disturbance and accessibility of UVs, scholars started implementing planting techniques to enhance the plant diversity of the site intentionally. Several researchers tested seeding techniques to increase plant diversity in UVs and confirmed the effectiveness of the broadcast seeding method [81,105,116], while others have proposed design frameworks for increasing plant diversity in UVs [92,96].

3.3.3. Landscape Factors Discussed in Reviewed Studies

Landscape factors refer to the influence of the surrounding environment on the plant diversity of UVs at the landscape scale, such as the composition of the landscape matrix, which is the majority of the surrounding landscape, around UVs and the level of urbanisation of the site location. These factors have been shown to affect seed dispersal and plant species composition in UVs, ultimately affecting plant beta diversity [13,93]. Based on the selected papers, the key landscape factors will include landscape connectivity (N = 21, 38%), landscape context or location (N = 12, 22%) and differentiation with other urban land (N = 3, 5%) (Figure 7).

From a landscape connectivity and isolation perspective, studies discovered that the similarity of plant composition decreased as the geographical distance between sites increased [64,83]. Similarly, the presence of nearby green spaces, such as urban parks and rivers, can also increase the species richness of various taxonomic plant groups inhabiting on site [86,90,117]. These green spaces provide a habitat for plants that can colonise UVs through the dispersal of seeds. Conversely, the presence of dwellings around UVs can isolate the site and act as a barrier to seed dispersal, and this can negatively affect the species composition of UVs [5,13,118]. To enhance the landscape connectivity of urban areas, Frazier and Bagchi-Sen (2015) [119] and Newman et al. [16] have developed models to identify UVs that can be used as a means of increasing connectivity across the entire landscape.

Several studies have found that the context of the surrounding landscape or location plays a crucial role in enhancing plant diversity in urban wastelands [14,85,88]. Specifically, increasing the distance from the city centre and decreasing the level of urbanisation around UVs has been shown to enhance plant diversity in these areas. However, Blouin, Pellerin and Poulin [5] discovered that the richness of non-native plant species was higher than that of native species at every level of local urbanisation intensity, with the highest richness found at a moderate level of urbanisation intensity. In addition, Brun and Di Pietro [64] found that plant communities are more similar in low-urbanisation contexts and more diversified in highly urbanised areas.

3.3.4. Methodological Factors Discussed in Reviewed Studies

Reviewed papers have explored the impact of UVs using different methods, such as the selection of different spatial scales, which include micro, local, and macro scales. Among these, the micro-scale—which refers to individual sites—was the most frequently studied (N = 36, 65%). The macro-scale, which examines UVs across entire urban areas, was the focus of 19 studies (35%). Only four studies (7%) investigated the local scale, which considers the city block or urban precinct as a unit of study. It is noteworthy that six studies compared UVs at two different spatial scales, while only one paper covered all three spatial scales simultaneously.

Figure 8 illustrates methods of analysis in studies that mostly employed case studies to investigate the plants on UVs and the factors that influence their diversity, with 46 studies (84%) using this approach. This can be further subdivided into vegetation sampling (N = 42, 76%), soil sampling (N = 7, 13%), and project design (N = 3, 5%). In addition, 13 articles (24%) used urban databases to obtain information on UVs, while only six articles (11%) employed experimental methods to explore the plant biodiversity of UVs. Three articles (5%) applied remote sensing techniques to obtain geospatial information from UVs, which include the use of software platforms to retrieve satellite imagery, such as Google Earth Pro version 7.1 (N = 2, 4%) and iTree software version 5.0 (N = 1, 2%). Furthermore, two studies (4%) applied desktop literature review as the main method and one article (2%) used social interviews.

Regarding data analysis methods, most studies (N = 42, 76%) utilised statistically based approaches, with analysis of variance (ANOVA) being the most commonly used technique (N = 20, 36%) as it quantifies the statistically significant differences between specific groups. Regression analysis (RA) was also widely applied (N = 13, 24%), as well as correspondence analysis (CA) (N = 6, 11%). Nine studies (16%) employed index models such as the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Built-up Index (NDBI) to evaluate vegetation and built-up land characteristics influencing UVs. For analysing collected data, ‘R’ (version 2.11.1-4.1.2) and ‘ArcGIS’ (version 10.2-10.6.1) were the most frequently used software, with 32 and 19 studies using them, respectively. Some of the literature references (N= 3, 5%) also utilised other analytical software, such as SPSS (version 16.0-23.0) and MATLAB R2017a.

4. Discussion

This systematic literature review was undertaken with the objective of evaluating the available global evidence pertaining to the factors responsible for plant diversity in UVs, considering the biophysical and temporal characteristics of the UVs and their landscape context. We have three main findings: the geographical specificity of current cultural approaches to urban voids, the significance and emphasis of a limited number of biophysical factors to study and understand plant biodiversity in urban voids, and finally, the role of scale and spatial factors in promoting or limiting biodiversity.

4.1. Cultural Bias in the Study of Urban Voids

Despite significant urbanisation in regions such as Africa and shrinking city trends in countries such as Japan, there is little research to describe the role that new urban voids have in these urban contexts and biomes. As noted by Rupprecht and Byrne [71], this study has identified and confirmed a noticeable bias in the literature on the plant diversity of UVs towards certain regions, particularly Europe, the USA, and China. The study on plant diversity of UVs has predominantly concentrated on the United States, wherein the number of UVs has increased due to the expansion of political borders in urban areas or the decline in population [120]. This trend has prompted increased research interest in understanding the value of novel ecosystems characterised by the presence of UVs [108]. Studies on UVs in the UK and other European countries are slightly different from those conducted in other regions, as they tend to focus more on brownfields, a specific type of UVs. The presence of potential pollution and contamination in brownfields has driven the research interest in understanding the role of soil conditions in influencing plant diversity in UVs. As such, soil analysis has become a central focus in many European studies on brownfields and plant diversity [121]. Asian countries, specifically China, have also shown a particular interest in brownfield sites with a history of coal mining. A range of research has explored the impact of site-specific biophysical factors, such as soil attributes and metal concentration, and restoration strategies [104,122,123,124].

It is also worth noting the contrasting development trajectories of UV research in Chinese and English literature. Over the last two decades, discussions on UVs and biodiversity have converged in English. In Chinese literature, these topics have remained relatively separate. Chinese landscape ecologists have tended to approach urban biodiversity primarily through macro-scale landscape analysis, studying how patterns and urbanisation affect ecological communities and their biodiversity [125,126,127]. Such research has highlighted how fragmented habitats impact biodiversity and has sought to promote connectivity strategies to address this problem [128,129,130]. Other Chinese studies have recognised ecosystem services within such spaces and suggested landscape security patterns for conservation [131,132]. However, Chinese studies have rarely examined specific UV types and their biodiversity and have instead focused on urban parks [133].

In China, UVs have been a topic of interest within urban planning research. More recently, they have been addressed by landscape architects who have expanded the focus on the topic to address design strategies [134,135,136,137]. In such domains, the ecological value of brownfields has been discussed [138,139], but biodiversity has seldom been specifically mentioned. The integration of ecological approaches with UV design has been stimulated in recent times by Kongjian Yu’s innovative design for Zhongshan Shipyard Park. Following this landmark project, other Chinese researchers and practitioners have increasingly examined spontaneous vegetation and biodiversity in industrial wastelands [140,141].

The geographical bias in UV research has significant implications for the development of context-specific conservation measures in regions with large urban populations, including India, South-east Asia, Africa, and South America, where rapid urbanisation and the loss of biodiversity are pressing concerns [142,143]. It is possible, however, that the literature on UVs in these regions may not be framed around plant diversity conservation but rather around food security. Therefore, a better understanding of local UVs could help devise strategies for preserving urban biodiversity in these areas, which require local knowledge to be effective.

4.2. Factors That Influence Plant Biodiversity in UVs

Based on the reviewed evidence, the factors that influence plant diversity in UVs can be classified into three high-level groups: biophysical factors, temporal factors, and landscape factors. Many studies addressed more than one type of factor, but only a small number considered all three categories simultaneously. For physical factors, our results are consistent with the findings of Bonthoux et al. [15] and Rupprecht et al. [111], agreeing that the majority of studies focused on the impact of site size, age, and land-use legacy, while the studies on the impact of site shape, microclimate, and ownership on UVs plant diversity were limited.

In terms of temporal factors, the reviewed evidence suggests that reducing the management frequency is important for maintaining UV biodiversity. For instance, changing mowing intervals or reducing them can not only improve site diversity and resource conservation but also help to maintain the natural site character, which is highly valued by local residents [111]. Besides, restricting residents’ access may also have a negative impact on UV plant diversity, as the diversity of formal and informal uses produces habitat diversity and local features that make UVs valuable for biodiversity [15]. This study further emphasises the importance of intervention strategies purposed for enhancing UV plant diversity, a point that has not been fully addressed in other previous reviews.

Regarding landscape factors, several studies have examined UVs in terms of their structural connectivity and isolation within the larger landscape scale, and it is worth noting that several research studies have indicated that the landscape context, especially the presence of green spaces, such as nearby UVs and urban parks, is positively associated with the species richness of plants. These green spaces serve as habitats for flora and fauna that can colonise UVs through the dispersal of spores or seeds or active movement. Conversely, the proportion of buildings surrounding the wasteland can act as a barrier to dispersal and may negatively impact the species composition of the UVs.

Only a limited number of articles explicitly mentioned multiple factors together (N = 2, 4%) or incorporated all three categories of factors (N = 4, 7%) in their analysis. This highlights the need for a more comprehensive and integrated approach in future research that considers the holistic study of the interplay between biophysical, landscape, and temporal factors to better understand their combinatory effects on plant diversity in UVs (note Figure 3).

4.3. Scale and Factors of Analysis

The scale of analysis plays a crucial role in identifying factors influencing plant diversity in UVs. Physical and temporal factors have been studied at the micro-level, focusing on the alpha diversity of plants, while spatial factors have been examined at the local and macro-level, emphasising the beta diversity of vegetation. The primary methods used to identify factors influencing plant diversity in UVs were case studies and statistical analysis. Case studies were prevalent in the literature due to the ability to conduct vegetation surveys at individual sites and gather quantitative plant data. However, there is no standardised case study method in this domain, and researchers often combine it with other techniques, such as literature reviews or statistical analyses, such as GIS. Although GIS-based methods were not commonly used in previous studies on informal green spaces [71,111], this systematic review reveals a growing trend in using GIS-based methods for analysing case studies of plant diversity in UVs.

4.4. Limitations of the Research and Future Studies

This review focuses on research between 2005 and 2022, with a particular focus on the differences and similarities between Chinese and English literature. Prior to this time period, different research clusters were located in Europe. If the search parameters were changed and earlier sources (prior to 2005) were included in the analysis, the picture would be different, and European research would be more dominant. As such, this limited time frame doesn’t provide a wide historical overview but rather a synchronic two-decade snapshot of research. More recent literature has also been considered through targeted critical review, but this has not been systematically integrated into the completed analyses.

The geographical bias noted above may, in part, be attributed to the limitations of this study, which only searched for studies published in English and Chinese, so studies published in other languages or available in other regional databases may be underrepresented. Nevertheless, papers published in Chinese only account for a small part of the studies conducted in China. This suggests that the comparatively low number of English publications on UV plant biodiversity may not solely result from missing non-English publications. Instead, it could indicate an actual gap in our knowledge about UV plant diversity in these countries. While this review, conducted in both English and Chinese, begins to address current research limitations, future reviews can target additional languages to clarify the issue of geographic bias.

5. Conclusions

This systematic literature review has identified the most relevant factors and dimensions of UV research in relation to biodiversity. Achieving a comprehensive understanding of the factors that affect plant diversity in UVs is crucial for effectively addressing urban sustainability, assessing biodiversity conditions, and implementing planting and design strategies. To this end, this systematic review sought to examine how authors worldwide have identified and measured these factors. The results indicate that the factors studied in most research can be broadly classified into three categories: biophysical, temporal, and landscape. However, we have noted that the majority of studies have only focused on one or two categories, and studies that comprehensively examined all three categories are rare. Additional exploration is required to discern how these categories can be effectively studied together and adapted to specific research objectives.

The need for localised knowledge of urban voids (UVs) is a significant gap in the current literature. The literature reviewed in this study indicates a lack of research on plant diversity in UVs outside of the United States and Europe, and international comparisons of UVs are rare. Furthermore, the lack of studies in many regions limits the potential for meta-analysis and cross-cultural studies. However, it is important to acknowledge that this study only examined articles written in English and Chinese. Future research could broaden the scope by including publications written in other languages in order to provide a more comprehensive understanding of UVs in different regions and promote global efforts for biodiversity conservation in urban areas.

Most studies have emphasised site-level natural characteristics as biophysical factors, site management as temporal factors, and landscape context as spatial factors. Future research should concentrate on refining and expanding the high-level categories presented in this review. For example, researchers could look deeper into the specific biophysical characteristics that impact plant diversity in UVs, including but not limited to land use legacy, hydrology, and microclimate. Furthermore, investigations into the temporal factors affecting plant diversity could examine how different human activities, such as dog-walking and urban agriculture in UVs, impact vegetation composition over time. Lastly, there is a need for more research at the macro level to examine how different landscape features, such as street trees, surrounding buildings, and anthropogenic activities, may influence plant diversity in UVs.

In addition, future research can focus on investigating the potential role of UVs in supporting plant species groups that have received limited attention, such as grassland and early succession forest species. Moreover, it would be beneficial to conduct long-term studies to investigate the influence of factors on plant diversity in UVs over extended periods. The identification and resolution of these critical gaps regarding UVs represents a fundamental step forward in achieving a more comprehensive and holistic understanding of their local features and, thus, their contributions to biodiversity, management, and integration into larger urban conservation strategies. This understanding is essential for effective conservation efforts in urban environments.