Between paper and plan: contrasting data on urban habitats in literature with planning documents

The most extensive habitat datasets found in the literature are phytosociological surveys. Over the last 100 years, a total of 148 publications containing phytosociological surveys as a form of documentation on urban habitats have been recorded in the studied cities (Figure 1). The number of these publications steadily increased until the second decade of the 21st century, reaching a maximum of 50 publications per decade. However, in the subsequent decade, it declined to 36 publications. Another type of habitat data derived from textual sources totalled 93 publications. The number has been steadily increasing since the 1990s, reaching 44 publications in the 2020s. The least represented type of data on habitats is maps, with a total of 37 sources. However, their number has been growing since the 1980s, reaching 13 publications in the 2020s.

Figure 1.

Number of publications containing data on urban habitats in respective years

Source: own elaboration

The number of habitats identified in individual cities varies considerably. The richest source is the Upper Silesian Urban Area, where 331 units were recorded. Meanwhile, the smallest is in Wałbrzych – 48 (Table 1). Among the cities distinguished by the highest number of identified habitats, in addition to the Upper Silesian Urban Area, are also Poznań and Warsaw. Here, 311 and 283 habitats were recorded, respectively. These values considerably exceed the average for all cities – 149.

The number of habitats specified in the planning documentation of cities is considerably lower than the number of identified habitats from the literature searches (Table 1). On average, 33.2% of habitats identified in research articles appear in planning documentation. This proportion only exceeds 90% in two cities, namely, Koszalin and Rybnik. For four cities, that is, Bydgoszcz, Poznań, Szczecin, and Tarnów, no habitat was recorded in the planning documentation. For over half of the cities, the number of habitats recorded in planning documents is under half that of the number identified based on literature.

All the habitat types identified in the cities belong to 29 classes according to plant community system (Table 2). The most common habitat types were Querco–Fagetea and Molinio–Arrhenatheretea, with each class containing 28 distinct habitats. These classes were also the most accurately represented in urban planning documents, with Querco–Fagetea appearing in 78.5% and Molinio–Arrhenatheretea in 64.3% of cases (Table 2). Conversely, habitat types from the classes Ammophilletea, Asteretea, Charetea, and Oxycocco–Sphagnetea were absent from the planning documentation.

The most common group mentioned in planning documents is forest habitats (Figure 2). They best reflect the actual state of knowledge, representing 62.4% of the habitats identified in the literature. Meanwhile, the synanthropic habitat group is the least frequently documented in planning documents. It is also the least represented, accounting for only 28.4% compared to the state identified in the literature.

Figure 2.

Habitat groups identified in the literature and included in planning documents, analyzing the period from 1980 to the present

Source: own elaboration

The amount of habitat data available in literature on Poland’s largest cities has been steadily increasing since the early 20th century. In the last decade, there has been a slight decrease in the proportion of phytosociological survey data. There has also been a slight increase in the proportion of maps. The most prominent increase has been observed in compilations, tables, and textual summaries. The number of cartographic and textual data is increasing the most rapidly. This is associated with the dissemination of conservation programmes based on formalized documentation and especially Natura 2000. Their introduction has resulted in the largest current database on nature across European countries (Mücher et al. 2009).

Studies containing phytosociological surveys typically focus on smaller parts of cities, often protected areas of less than a square kilometre. Their number increased until 2010, after which a decline occurred. It is challenging to definitively determine whether this is a lasting trend or a short term dip. Additionally, studies encompassing habitats throughout the city are extremely rare. This is due to the time-consuming nature and high cost of such studies (Jarvis & Young 2005). In Poland, there are relatively few examples of such studies, including Białystok, Kraków, and Warsaw (Sudnik-Wójcikowska 1987; Wołkowycki 2019; Zając et al. 2006).

Habitat recognition in individual cities is uneven and any cities have limited documentation. The degree of habitat recognition in urban areas depends on the natural characteristics of each city and the focus and capabilities of local research institutions. The best-researched ones are those with higher education institutions and research institutes with a long tradition in phytosociology. Cities such as Poznań, Warsaw, the Upper Silesian Urban Area, Białystok, Lublin, and Bydgoszcz stand out here (Table 1). For example, Warsaw benefits from comprehensive cartographic data, facilitating accurate habitat mapping. In contrast, Poznań has detailed information about specific habitats, offering in-depth insights into their ecology. This has highlighted how local research priorities and resources influence the quality and extent of habitat data available for urban planning and conservation.

Our study has indicated that disparities are not only present between habitats identified in literature and those documented in planning documents. These disparities are also evident in the inclusion of different habitat groups in planning documentation, particularly for forest habitats. Their frequent inclusion in plans is because of their historical connection with the urban fabric (Forrest & Konijnendijk 2005). Due to urban development, forests and woodlands have often been transformed into parks or urban forests which have high cultural and historical importance for local communities (McBride 2017). Another factor is the role of forests in maintaining biodiversity and protecting natural resources, making them key to city ecological sustainability. Spatial planning often prioritizes the protection and integration of forested areas into the urban fabric (Muller et al. 2010). Urban forests also perform essential ecosystem functions, such as air purification, microclimate regulation, and water retention, further enhancing their value in the eyes of urban planners (Nowak 2006). All these factors make forest habitats the most frequently documented and included in urban planning documents. This has highlighted their position in the context of city spatial management and nature conservation.

Synanthropic habitats are one of the most transformed and degraded areas within city boundaries (Kowarik 2011; Manyani et al. 2021). They mainly contain plants that have adapted to unfavourable urban conditions. However, due to their frequent location in dense urban fabric, they can also become habitats for many invasive species (Aronson et al. 2017; Culley et al. 2022). This combination of native and invasive species can lead to the formation of new, stable communities unknown to phytosociologists/botanists. These communities align with the concept of so-called novel ecosystems (Hobbs et al. 2013; Kowarik 2011). In urban areas, novel ecosystems can occupy up to half of the entire city, encompassing wastelands, city parks, and neighbourhood greenery (Kowarik 2011; Rupprecht & Byrne 2014).

Our study has shown that synanthropic habitats are the most overlooked group in planning documents. This is caused by several interconnected factors. For a long time, synanthropic city habitats were only of interest to a small group of researchers (Yeremenko 2019). Historically, these habitats were viewed as unstable and transitional, with the assumption that they would eventually develop to resemble natural habitats. The low recognition of these areas resulted in white spots on vegetation maps of cities. These habitats often occupy degraded, postindustrial areas, leading them to be treated not as suitable green spaces, but rather as vacant lots available for further urban development. This has resulted in a systematic overlooking of them as areas requiring recognition as green spaces, instead focusing planners on developing these areas. Moreover, there has been a longstanding belief that residents negatively assess these areas in terms of aesthetics and their usefulness as urban green spaces (Nassauer 1995).

Synanthropic habitats provide numerous ecosystem services in the urban fabric, such as air purification, water management or recreation, and their value is relatively high (Luo & Patuano 2023). Wastelands seem to be a crucial group of these ecosystems. Due to the almost total lack of human interference in these areas, vegetation undergoes natural succession processes (Bonthoux et al. 2014), where vegetation develops spontaneously (Schadek et al. 2009; Yuan et al. 2023). Therefore, a single area may exhibit high habitat diversity at different successional stages with different floristic compositions. Therefore, wastelands become specific urban biodiversity hotspots. Considering their high diversity value, it is disconcerting that they are so minimally considered in planning documents, especially in the face of trends such as resilience and sustainable cities. This is alarming, especially considering that, on average, they can comprise up to 15% of a city’s total green space (Sikorska et al. 2020).

There is a considerable need for a better understanding of the role of habitats in urban areas and their more precise identification in space (Table 1). With the increasing use of remote sensing in urban habitat mapping (Neyns & Canters 2022; Yan et al. 2018), the potential of such tools is growing. Remote sensing data are successfully used to assess habitat quality (Lakes & Kim 2012). Their usefulness can be high considering the high costs associated with habitat mapping using traditional methods. Along with geographic information systems, these technologies have provided new possibilities for accurate and rapid mapping. Further, the obtained data are more reliable and easier to update than paper records. Given that remote sensing is based on ground-validated data, there is a risk of omitting unknown data. Mapping techniques are often adapted to individual types of greenery (Neyns & Canters 2022). Therefore, it is essential to coordinate the collection of field data with remote sensing techniques to maximize the remote sensing accuracy (Chan et al. 2021).