1 Introduction

Rural settlements in the Czech Republic form a dense network of settlements providing housing to about one-third of the country's population. This network includes settlements that were in many cases founded as early as the Middle Ages (Máčel, 1954, 1955; Tůma & Pánek, 2009). The administrative division defines settlements (municipalities) that do not reach the limit of 3,000 population as rural settlements. They are also characterised by an immediate connection with the open landscape and a gradual transition in which the green space located in the backyards and the outer belt of the open landscape surrounding the settlement plays a key role (Holland & Risser, 1991; Hufkens et al., 2008; Mareček, 2008; Marfo et al., 2018; McKinney, 2006).

The rural landscape arises as a combination of natural and cultural elements (Agnoletti, 2014; Hunt et al., 2002; Ryan, 2002; Thorbeck, 2013) and results from the land management by many generations of farmers (Sklenička, 2002). Until about a century ago, the rural landscape represented an important part of the daily life of the inhabitants of country areas, providing subsistence through agriculture (Perlín et al., 2010). Over the past century, however, rural landscape has experienced many changes, in particular a sharp decline in diversity, e.g. due to the consolidation of small fields, destruction of baulks and copses, sunken lanes, etc. (Sklenicka et al., 2014a; Šťastná et al., 2018), which led to the simplification of the landscape structure and loss of valuable ecosystems (Sklenička, 2003; Št’astná et al., 2015). This decline in diversity was caused predominantly by the intensification and mechanisation of socialist agriculture and its collectivisation (Sklenička, 2002). After the end of the socialist regime and its central planning in 1989, rural settlements in the hinterland of large cities experienced (along with the transition to the market economy) dramatic development (i.e. suburbanisation) (Bičík et al., 2010; Maier, 1998; OECD, 2017; Šťastná et al., 2018) accelerated by globalisation (Kocur-Bera & Pszenny, 2020). In addition, the inhabitants of rural settlements gradually lost their dependence on the landscape surrounding their settlements. The population of rural settlements in the hinterland of large cities are mostly no longer employed in agriculture, which is another cause for the loss of the rural character of these settlements (Perlín et al., 2010).

The rapid development of rural settlements has raised concerns about the loss of their identity (Baše, 2004; Foley & Scott, 2014; Kocur-Bera & Pszenny, 2020; Taylor, 2011) and disruption of the landscape character consisting, among others, in the harmonious and organic connection of the rural settlement with the open landscape (Löw & Míchal, 2003). Nowadays, many rural settlements resemble (through the density and character of the constructions and buildings) rather urban settlements than villages (Heyer, 1990). This is especially due to the building expansion at the expense of quality agricultural land (Baše, 2004; Kocur-Bera & Pszenny, 2020; Skaloš et al., 2012; Titzenthalerová, 2012) and backyard space (Psotová, 2008). Settlements then tend to grow, merge with other settlements, and lose their character and identity, becoming only a part of the metropolitan area (Baše, 2004). The rural landscape thus turns into the suburban landscape (Forman & Gordon, 1993).

In view of the global population growth (Field et al., 2012), the trend of urban agglomeration can be expected to continue in the Czech Republic as well (OECD, 2017). This is also associated with the increasing representation of the built-up and paved areas at the expense of water-absorbing areas. These changes reduce the retention capacity of the landscape, which in turn leads to a drop in the groundwater levels. Built-up surfaces further lead to the formation of so-called urban heat islands (UHI) (Armson et al., 2012; Oke, 1982), characterised by the increase in the temperature compared to the surroundings of the settlement. This higher temperature can, among other things, also negatively affect the population's health (Arsenović et al., 2019; Clarke, 1972; Shimoda, 2003; Středová et al., 2015) as well as the energy consumption due to air-conditioning (Alavipanah et al., 2015; Oke, 1982).

This paper aims to contribute to the understanding of the dynamics of the development of the periphery of rural settlements in the hinterland of large cities. The knowledge of the historical development and the causes leading to the current rural form can serve as a basis for understanding the processes that have contributed to the creation of the present-day landscape structure and the condition of its environment. Finally, the paper aims to:

  • analyse the evolution of three suburbanised settlements from the perspective of the representation of water-absorbing surfaces and their suitability for planting taller vegetation in public and private spaces with an emphasis on the settlement margins over the period 1846–2021

  • propose recommendations on measures that could help avoid negative impacts of the external factors on rural settlements, stabilise and enhance rural character, and avoid negative impacts on the open landscape.

2 Materials and methods

In this study, the margins of three randomly selected rapidly developing rural settlements in the hinterland of the metropolitan region were qualitatively analysed. Creative interpretation requires knowledge of the genesis and changes of settlement margins in relation to the open landscape, and so these edges were analysed from the mid-nineteenth century, when the first reliable map evidence appears, to the present day.

Here, a detailed analysis was performed to reveal the trends of rural settlements and their margins in the metropolitan area. Despite being just a case study, its results can be generalised as an example of the evolution of the use of rural settlement margins, which have been strongly influenced by the building expansion associated with suburbanisation.

3 The study area

The hinterland of Prague, the capital of the Czech Republic, Central Europe, was chosen as the study area, from which three rural settlements (Chýně, Květnice and Nehvizdy) were randomly selected. The relatively small number of settlements was selected intentionally, due to the time-consuming nature of the analysis at individual time points. The extents of these three sites were analysed at three-time points (see Fig. 1). Dark grey indicates the historical core of the sites in 1845. The medium grey (hereinafter referred to as Development 1845–1951) shows the area newly developed area between 1845 and 1951 and, analogically, the light grey (referred to as Development 1951–2021) covers the areas developed between 1951 and 2021.

Fig. 1
figure 1

Location of the samples site in the hinterland of Prague, Czech Republic, Central Europe, and extent of the sites at individual time points; Dark grey: historical core (1845); medium grey: area newly developed between 1845 and 1951; light grey: areas developed between 1951 and 2021

Full size image

4 Data sources, collection, and verification

In all, three-time points were analysed, each represented by a different type of data. The oldest time horizon was recorded using cadastral maps, the middle one using aerial photographs and land cadastre records, and the youngest one using cadastral maps and satellite images. The studied time points are, among other things, interesting in the fact that they represent significant points of political and economic changes that had a significant impact on the character of settlements and open landscape (Kupka, 2010; Kupková et al., 2021; Maier, 1998; Sklenička, 2002; Sklenicka et al., 2014b).

4.1 Mid-nineteenth century

This time point characterises the structure of the areas at the beginning of the modern (industrialised) era in Czechia. In other words, this is a time point when the even older character of the landscape was maintained as the new requirements on the land use and structure (e.g. the development augmented by railroads) have not been implemented yet. The first source used for analysis, map sheets of the Stable Cadastre from the first half of the nineteenth century, represent a relatively accurate record of the built-up areas and open landscape surrounding it. These maps have been successfully used previously, e.g. in Land Use and Land Cover (LULC) analyses (Bičík, 2020; Bičík et al., 2001, 2010; Kupková et al., 2021; Skaloš et al., 2011, 2012).

To be able to evaluate the changes in the land-use boundaries between the historical and present-day settlement, the maps were georeferenced in ArcMap 10.7 software using several ground control points with known locations on maps from all time periods.

4.2 Mid-twentieth century

The second time point represents one of the most important milestones in the history of the Czech cultural landscape (Kupka, 2010). At that time, agricultural collectivisation was at its peak, bringing systematic changes to the landscape character (Bičík et al., 2010; Lipsky, 1995; Sklenicka et al., 2014b). Black and white orthophoto images from 1953 were used to capture the landscape cover, and the Land Cadastre from the same year was used to identify plot boundaries and was georeferenced.

4.3 Present day

In the late 1990s and the beginning of the new millennium, the study area saw an explosive suburbanisation wave resulting from the change in the political system (Maier, 1998; Ouředníček, 2007; Sýkora & Mulíček, 2014) and new requirements on the land use (Doucha, 2002). Several data sources were used to identify the current LULC. One of those was aerial imagery, including near-infrared radiation (NIR with wavelengths from 780 to 2500 nm). Artificial intelligence was trained and subsequently used to distinguish water-absorbing and non-water-absorbing areas (see below). Current data from the land cadastre registry were used to identify plot boundaries and locate buildings.)

5 Data processing and analysis

For the purpose of the analysis, the settlement areas were divided into water-absorbing (permeable), and paved and built-up (non-water-absorbing) areas. The permeable areas represented mainly gardens, orchards and parks. Paved areas included roads, paths, sidewalks or courtyards, and built-up areas (houses, garages, outbuildings, etc.). In addition, the plot sizes were analysed and special attention was paid to the differences between private (inaccessible) and public (accessible) areas at all three time points.

In order to be able to analyse the development of the settlements over time and, additionally, the changes in their relationship to the open landscape, it was necessary to define the margins of these settlements. For all periods, the same approach was used—the margins were defined as an envelope enclosing the built-up areas including the backyards (gardens) of private properties, businesses, and public spaces.

Using the thematic layers prepared for individual settlements and time points as described above, GIS analyses (ArcMap 10.7 and ArcGIS Pro) were used to determine the representation of the individual surface classes (water absorbing, non-water absorbing; the latter was further divided into built-up and paved areas).

5.1 Processing of the present-day data

For present-day data, automated classification of the surfaces into water-permeable and non-water permeable was performed using airborne imagery (visible and near-infrared (NIR), combined into colour-infrared (CIR) imagery). The visible spectrum imagery was obtained from the State Administration of Land Surveying and Cadastre; the NIR data were supplied by the Forest Management Institute of the Czech Republic. The acquired imagery was first merged into a single layer, and subsequently, the NIR segmentation for the needs of the analysis was performed, which led to the unification of pixels of similar colour into individual vector surfaces, therefore converting the raster image into evenly-coloured polygonal segments. The Spectral Details value, representing the degree of differentiation between similar colours in the segmented image and, thus, determining the number of segment classes, was set to 16. The Spatial Detail value, which sets the level of spatial detail of the segmentation process, was set to 0, which ensures smoothing the noise out from the output. The Minimum Segment value, which controls the minimum size of segments, was set to 20 pixels, resulting in a good trade-off between ignoring pattern interruptions (e.g. chimney on a roof) and maintaining overall precision.

5.2 Machine learning analysis

The individual vector surfaces obtained in the previous step were subsequently classified into water-absorbing (WA) and non-water-absorbing (NWA) surfaces using the geoprocessing tool Support Vector Machine Classifier (ArcGIS Pro, Image Analyst plugin), which can classify large volumes of data through machine learning on user-supplied training data. Thus, it was necessary to create a training database of approximately 100 samples from the categories of red roofs, dark roofs, light roofs, driveways, and roads for NWA surfaces, and bare ground and greenery for WA surfaces. These training data samples were provided across all investigated settlements. Samples were either defined using a hand-drawn polygon, or using the resulting segment from the previous segmentation process (see Fig. 2).

Fig. 2
figure 2

CIR orthophoto image showing data on a sample of the Nehvizdy settlement

Full size image

The sample database was then used by the Support Vector Machine Classifier machine learning algorithm in order to train a classification method and this trained algorithm was subsequently employed to classify the segmented CIR image into predefined categories. The classification result depends, in particular, on Segment Attributes describing different colour characteristics and features of the segmented shapes. Setting these attributes further enhanced the accuracy of classification. As the classification process was completed, each segment was assigned into one of the seven categories (see Fig. 2) used by the previous sampling process. For the purposes of this paper, these categories were further aggregated into WA and NWA surfaces, reducing the number of categories from seven to two.

5.3 Accuracy of the classification process

In order to be able to assess the accuracy of the classification process, accuracy assessment points were randomly created using the geoprocessing tool Create Accuracy Assessment Points (ArcGIS Pro, Image Analyst plugin). This process has been set to create 100 points randomly distributed in the study area, while making sure that half of those points are located at WA and the other half at NWA surfaces.

Subsequently, the accuracy assessment points have been visually compared to the basic orthophoto (no infrared bands, no segmentation, no classification), manually and individually assigned the absorbing/water non-absorbing attribute, and recorded in an attribute table. This table served as an input for yet another geoprocessing tool—Compute Confusion Matrix (ArcGIS Pro, Image Analyst plugin). The output of this tool, a confusion matrix, represents the accuracy of the classification process. The following values (ranging from 0 to 1) within the confusion matrix are the most important: individual accuracies of both WA and NWA detection and the Kappa value, which represents the overall accuracy of the classification process. The classification accuracies in this paper were: the accuracy of WA detection 0.96, NWA detection 0.98, and Kappa value 0.94. The Kappa value above 0.9 indicates a good reliability of the classification process. It can be also observed that the detection of NWA surfaces was more accurate than the detection of WA surfaces.

5.4 Identification of built-up areas and of areas suitable for planting taller vegetation

The resulting classification was subsequently refined and verified based on cadastral data and orthophoto maps, correcting any discrepancies between reality and cadastral maps (e.g. new buildings not yet included in the CIR imagery.

Further analysis of the quality of WA surfaces was performed to identify areas suitable for taller vegetation and those with the potential for enhancing biodiversity. This was done by excluding too narrow or too small segments of WA surfaces. Any WA area that is at least two meters from the border of NWA and at least 2 m from the plot borders was considered eligible to provide the required biological and retention functions for “taller vegetation” of the area. The remaining WA areas not meeting the aforementioned criteria were classified as “lawns”. After classification, we see that such areas are much more represented at the margins than in the central part of the settlement (see Fig. 3).

Fig. 3
figure 3

Orthophoto map (1) supplemented with NIR data (2) and the resulting classified layer (3); light green: lawns; dark green: soil surfaces suitable for taller vegetation; grey: paved surfaces; dark grey: built-up areas)

Full size image

Furthermore, the representation of the surface types in private plots neighbouring on the margins of the settlement was analysed. Three 15-m-wide bands were considered (A, B, and C, with the band A being the outermost, directly neighbouring on the settlement margin, and band C being the most centrally located. In this way, the 45-m inner margin of the settlement was evaluated at each time point. As described above, the areas were categorised into WA and NWA surfaces, and these were further subcategorised into built-up/paved areas (NWA) and lawns/surfaces suitable for taller vegetation (WA).

In addition, a 50-m belt of the open landscape neighbouring on the built-up plots was analysed based on the orthophoto maps and cadastral data, classifying the surfaces into fields, meadows, water areas, line constructions, baulks, and forests; a further subclassification into dirt roads and paved roads/railroads was also performed.

6 Results

The results show dynamic changes in the relationship between rural settlements and the open landscape. The tendency to the development of sharper, less gradual, borders between the settlement and landscape was observed in all three study areas. The main findings are summarised below.

6.1 Change in plot sizes

The trends regarding plot size were similar in all three study sites; for this reason, the figures below contain aggregate data for all three sites. It is obvious that in the historical core of the settlement, a gradual shrinking of plots over time can be observed, which results from their gradual division. Over the nearly 200 years, the size of the plots in the original (historical) core has thus been reduced by about half (see Fig. 4). This trend is further strengthened by their central location, which is under greater pressure during redevelopment, i.e. the water WA areas are converted into NWA areas (see Table 1). This trend is most evident in the plots located on the margins of the historical core where the original large gardens have been subdivided. These original gardens were in the second (1950s), and especially the third (twenty-first century) time points largely converted into residential areas, which resulted in the increase in the proportion of WA areas (a 10–20% loss in the core locations). The trend of plot size reduction was especially apparent between the first two time points; between 1951 and 1921, the reduction in plot sizes was not so notable. The vast majority of plots do not exceed 1000 m2 in size. The sizes of plots in the modern developments (1951–2021) are significantly smaller compared to both the historical core and 1951 areas. These private plots are not only smaller (median area of 521 m2; similar results were reported e.g. by (Grose, 2009)), but the variability in size as well as in the representation of WA and NWA areas is much lower than in the older parts of the settlement, see Fig. 5.

Fig. 4
figure 4

Evolution of the sizes of private plots from the nineteenth to the twenty-first century (combined representation from all three sites)

Full size image
Table 1 Representation of the water-absorbing, built-up and paved areas in individual study sites and time points
Full size table
Fig. 5
figure 5

Historical development of the representation of the WA (green), paved (grey), and built-up (black) areas in the public space of the three sites (combined representation from all three sites)

Full size image

6.2 The growth of the rural settlement and the evolution of the three principal components

In all three study areas, a trend of the increasing proportion of areas of paved and built-up surfaces over time at the expense of WA surfaces can be observed. In the historical core of the first site, the increase in the representation of the built-up and paved areas was observed between 1845 and 1951, whereas by 2021 there was a slight decrease in their representation; the opposite trend was, obviously, observed in WA areas at this site.

In all three study sites, a generally similar tendency of the increasing representation of built-up and paved areas at the expense of WA areas over time can be observed.

A more detailed look reveals a difference in the representation of WA and NWA surfaces between the public and private spaces.

In the case of public areas (Fig. 5), the relatively high present-day proportion (42%) of the WA component is partially due to areas that cannot be otherwise used. These are, therefore, not areas deliberately planned as urban green space; rather these areas are incidentally unsuitable for development. In most cases, there are just several areas that are relatively large, typically in inaccessible or otherwise unsuitable locations. Also, the public area of WA surfaces (representing greenery) in the historical core (largely consisting of the village square and adjacent streets in 1845) increased over time (from 11 to 35%). This is mainly caused by the improved differentiation between paved surfaces and greenery and by the development of parks in the historical core, sometimes even at the expense of original buildings.

In the new developments (i.e. the 1951 development and, especially, the 2021 development), the public space in the residential zones is predominantly formed by streets and greenery is omitted. Greenery in such zones is, therefore, present predominantly in the private plots; however, as obvious from Fig. 6, these plots do not have much space suitable for planting taller vegetation and most such spaces only serve mainly as lawns, with the occasional shrubs or thujas. Also, the apparently high percentage of the WA area in the public space in the newest developments (42%) must be interpreted with caution. It must be considered that the public space in these developments constitutes only a small fraction of these areas (less than 1/3 of the 1951–2021 development), and if an area unsuitable for construction activities is present within such development (i.e. steep slopes), it artificially inflates the relative representation of greenery in the public space.

Fig. 6
figure 6

Development of the classified surfaces in three 15 m wide belts on private plots at the settlement margins (%)

Full size image

6.3 Backyards

Historically, both in 1848 and 1951, the representation of greenery was decreasing in the direction towards the centre (the highest representation was in the outermost belt A, lowest in the innermost belt C). In 2021, however, the representation of greenery in belt B (15–30 m from the margin) is approx. the same level as in belt C (52% and 54%, respectively); this goes hand-in-hand with the almost 50% representation of non-water-absorbing surfaces (buildings and paved areas) in both bands. This trend is likely largely caused by the decreasing size of the private plots. The gradual transition between the open landscape and built-up/paved areas is therefore largely lost.

At the same time, the share of WA areas suitable for tall vegetation decreases (for example, from 56% in 1845 to 32% in 2021 in the outer belt, see Fig. 6; the same can be observed in the remaining belts as well).

In addition, within the NWA areas, an increase in the proportion of built-up surfaces relative to paved surfaces can be also observed, which is, again, likely caused by the decreasing size of individual plots (Fig. 7).

Fig. 7
figure 7

Evolution of the representation of paved, built-up and water-absorbing areas in plot sizes of varying sizes

Full size image

6.4 The proportion of the classified surfaces in plots of different sizes

In view of the aforementioned influence of the plot size on the representation of individual surfaces, an additional analysis taking into account plot sizes was performed. As the plot sizes evolved over time, the sizes in each period were classified into 7 quantiles.

Figure 8 shows the representation of greenery, built-up and paved areas in thus categorised plots. This graph reveals several trends.

Fig. 8
figure 8

Evolution of the outer margin of the settlements—land use/cover types

Full size image

When comparing the settlements at all three time points, it is obvious that the WA areas were more widely represented in the nineteenth century; then, these surfaces also served as a source of food (especially orchards or pastures for cattle). On the smallest plots, however, greenery was represented only to a very small degree.

In the twentieth century, a paradigm shift occurred and the association between the growth of the plot. The reason can be found in the fact that the production of food ceased to be the primary function of gardens as food could be easily procured. This was especially true for the rural population in the vicinity of large cities who often started to commute for work into the cities where they also kept purchasing food. In this way, rural settlements of predominantly rural character turned into so-called post-agrarian settlements. At the same time, as shown in Fig. 4, the representation of large plots, relicts from the nineteenth century, decreased. The gardens and their size thus lost their function from the past.

In the twenty-first century, the importance of the WA areas started to increase again, although the primary purpose of gardens turned into recreational use. Even in the smallest gardens, a certain balance between the built-up and permeable parts of the plot can be observed, which is likely largely due to the urban plan limitations prescribing the amount of WA areas that must remain in the plot. The WA component becomes an equal partner to the NWA one, but only in the case of smaller plots. For larger ones, interestingly, the built-up area increases at the expense of the WA areas. This is likely caused by the lifestyle and the fact that the large plots in the new developments are typically owned by wealthy individuals who use the available space for luxurious constructions (such as swimming pools or private sports fields). These new uses can be found quite frequently and represent new trends compared to the past.

6.5 Outer belt

Furthermore, an analysis of the immediate neighbourhood of the settlement was analysed. Below, a combined result describing all three sites at individual time points is presented (Fig. 8). The representation of the landscape types directly neighbouring on the settlement is influenced by the fact that the settlement growth brings the new margins of the settlement deep into the agricultural landscape, which is used to be relatively distant from the settlement core in the past.

The investigation of the neighbourhood of the analysed nineteenth-century rural settlements shows that roughly half of such immediate surroundings (outer belt round directly neighbouring on the Belt A on the outside) consisted in approx. 50% of agricultural land, roughly 18% of which was separated by a linear element (road or baulk). The 1951 data already displayed a change of this, with the loss of baulks and the intensification of agriculture, which is also in agreement with other studies (e.g. Bičík et al., 2001; Sklenicka et al., 2014b).

Paradoxically, the greater presence of the non-agricultural portion of the outer margins of the twenty-first-century developments has been caused by the extensive settlement development that has cut into the surrounding agricultural landscape, thus moving the fringes closer to the natural barriers, such as streams and forests. Such institutionally protected environments such as streams and their immediate surroundings (as well as forests, etc.) benefit the settlement by providing recreational opportunities and reducing negative impacts from the agricultural environment (such as erosion and dust). On the other hand, however, the settlement can negatively affect such natural barriers (e.g. noise, light pollution, etc.). However, not all such areas represent greenery of the forest type; often, these include unused areas waiting for a change in the urban plan. They are often overgrown with successional and emergent vegetation, a trend that has been increasingly occurring in the last few decades.

When analysing the relationships between the outer belt and backyards, the evolution over time is obvious. In the nineteenth century, the outer belt of the backyard zone predominantly consisted of WA surfaces (75%, of which 56% were suitable for taller vegetation), which formed a gradual transition, especially between the arable land and the rural settlement. Although the total greenery representation in the backyards has not changed much till present (from 75 to 68%), the representation of the surfaces suitable for planting trees or other taller vegetation dropped much more significantly (from 56 to 32% in zone A, from 53 to 26% in Zone B), see Fig. 6.

7 Discussion

The presented analysis of the margins of rural settlements in the hinterland of a large city shows the gradual change and loss of features characteristic of rural settlements in the Czech Republic. Excessive change of the original settlement and of its relationship to the open landscape into its current form may negatively impact the settlement, its inhabitants, and the adjacent open landscape, together with a decline in biodiversity, due to the synergistic effect of external influences. This negative impact was created by building expansion at the expense of quality agricultural land (Baše, 2004; Kocur-Bera & Pszenny, 2020; Skaloš et al., 2012; Titzenthalerová, 2012) and backyard space (Psotová, 2008). Such settlements then tend to grow and merge into the suburban landscape (Forman & Gordon, 1993).

7.1 Building density

The lower plot sizes and their higher density in the newest parts of the settlements compared to the original development result in less vegetation in these locations (Fig. 6). As a consequence, the water retention capacity is reduced and the reflectivity from surfaces increased, thus promoting the development of local heat islands (Bao et al., 2019; Barthel et al., 2017; Ramamurthy et al., 2017; Wang et al., 2019).

The urban structure of the new developments is, therefore, much different from the character of the historic core (see Fig. 9 for illustration). The new developments contain fewer WA surfaces potentially supporting also planting of taller vegetation such as (fruit) trees as the small plots cannot support this due to spatial constraints. The WA surfaces there often consist of narrow strips around built-up areas where planting taller vegetation is not feasible. In this respect, rural settlements are beginning to resemble the urban environment and lose the original attributes that have formed their character for many centuries.

Fig. 9
figure 9

Illustration of the permeability in the historical core built before 1845 (left) and in novel developments built at the beginning of the twenty-first century (right). Upper images represent the classified layer, the bottom images the orthophoto map. Legend for the classified layer: light green: lawns; dark green: soil surfaces suitable for taller vegetation; grey: paved surfaces; dark grey: built-up areas)

Full size image

7.2 Sharp transition

The aforementioned building development on smaller plots brings the built-up (and paved) areas closer to the open (agricultural) landscape (Fig. 6). New developments no longer contain as much WA area as in the past; this is especially true for surfaces suitable for planting taller vegetation (although the total WA area that must remain in the plot after construction is typically prescribed, such area is typically formed to a large degree by lawns). This is due to the trend of the times where, for example, people no longer need so much space for subsistence. At the same time, the price per developable area is too high, making it economically unviable to create large gardens the area of which could be used for planting taller vegetation capable of forming a dense canopy. Research carried out on this issue has previously highlighted a general trend towards smaller plot sizes, resulting in the area of the house itself occupying a significant proportion of the plots (Hall, 2019; Muminovic & Caton, 2018). This is consistent with the results of the presented study showing that the opportunity to plant taller vegetation is limited on current plots, which results in a decline in the amount of this type of vegetation in the peripheral parts of the settlements, thus removing the transition zone where tall vegetation used to be abundant. In effect, the relationships of the settlements with the surrounding open landscape are altered (Ryan, 2006), which leads to the “estrangement” of these settlements from the surrounding open landscape.

This may negatively influence the settlement and its inhabitants by external factors such as erosion and the associated soil degradation and drift (Burel & Baudry, 1995; Kristensen & Caspersen, 2002; Löw & Míchal, 2003; Sádlo, 2008) as well as by effects directly associated with farming in the area neighbouring with the settlement (noise, dust, odour, etc.).

The absence of this green belt results in high contrast between settlement and the surrounding open landscape, which can (among other things) lead to the loss of habitat for animal and plant species. The sharp transition between the open landscape and the backyards (Prudký, 2008) is becoming a common part of suburbanised settlements (Titzenthalerová, 2012), thus disrupting their character as well as the character of the landscape, which is used to have a gradual transition between the backyards and open landscape (Fig. 6).

The expansion of suburban settlements in the hinterland of large cities is sometimes so rapid that individual, originally isolated, settlements merge. This only reinforces the tendency to increase the proportion of the built-up, NWA surfaces and, among other things, supports the increase in the surface temperature (Fig. 10).

Fig. 10
figure 10

A typical example of a settlement margin in a new development where the built-up area directly borders on the open landscape, without the belt of taller vegetation that used to be a typical attribute of rural settlements

Full size image

Such gradual processes can act synergistically, increasing the potential risk of negative impacts of adverse weather and climate phenomena, such as longer and more intense heat waves and droughts (Barthel et al., 2017; Fischer & Schär, 2010; Tuholske et al., 2021).

7.3 Traditional rural character as a possible solution to the situation

The significant increase in the proportion of built-up and paved areas in rural settlements in metropolitan areas, together with the loss of greenery, calls for a comprehensive approach to this problem and to the future development of these settlements. The dynamic changes in the character of rural settlements are caused, among other things, by the fragmentation of the settlements into parts with different architectural forms.

Several recommendations can be derived from our study. The first one is to stabilise existing areas of the green spaces. This is, of course, not a new idea as it has been recommended elsewhere in the literature. Incorporating vegetation into the built-up environment generally has a positive impact on reducing temperatures (Armson et al., 2012; Field et al., 2012; Gill et al., 2007), as well as, e.g., on energy consumption (caused by the reduced need for air-condition) and CO2 emissions (Alavipanah et al., 2015). The presence of greenery also promotes other characteristics (aesthetics, psychology) (Finlay et al., 2015; Johansson et al., 2016; Ong, 2003). Thus, with respect to the type of settlement and its location, it is proposed to stabilise the remnants of taller vegetation that is found in the backyards of the historic cores of these settlements.

Secondly, this principle of planting taller greenery should also be applied to the outer belts of the current extent of the settlement. Such a green belt would create an environment enhancing biodiversity and serve many other purposes (Kowarik, 2019); among others, they provide a windbreak, thus preventing dust from surrounding fields to be carried into the settlements, a barrier against soil carried into the settlement from the fields during rains, etc. (Derkzen et al., 2015; McKinney, 2006). It could also have recreational use—such belts could be used for public footpaths, etc. Tall vegetation could, when employed in a sensitive manner, also contribute to screening out the negative aesthetics of suburban typified architecture that highly contrasts with the traditional regional architecture. Such visual screening could enhance the silhouette of the rural settlement (Prudký, 2008) as well as improve its structure and articulation. Results of the analysis of the outer belt and backyards also imply that the smaller plots and associated lower representation of WA surfaces that are suitable for planting taller vegetation largely prevent its placement in the backyards of private gardens. However, the optimal ecotone width and usage should be a matter of further investigation.

8 Conclusion

The presented paper analysed the nearly 200-year evolution of environmental changes in the margins between rural settlements in the hinterland of large cities and the surrounding open landscape. Attention was paid to the changes in land cover, the density (i.e. plot sizes) of developments in the peripheral areas of the settlements and the character of the transition of rural settlement to the open landscape.

The original historic cores of the rural settlements contain, compared with the newly developed sites, more water-absorbing (WA) areas suitable for planting (especially taller) greenery. The building boom that has occurred in the first decades of the twenty-first century largely destroyed the original margins of these original rural settlements (gardens and orchards with tall vegetation providing a gradual transition between the settlement and surrounding landscape). This results in the deterioration of the original characteristics of the rural settlements and a decline in the quality of the rural environment.

The rural settlements that were in the past formed in close relation to the open landscape have now turned their back on the open landscape, closing themselves off from it, and at the same time, eliminating the water-absorbing areas on which tall vegetation could be planted.

The recommendations resulting from this analysis, which were outlined at the end of Discussion, can be of benefit to spatial and landscape planners, as well as to the representatives of the municipalities of such settlements, and serve as a basis for the preparation of documents for the strategic development of rural settlements in the hinterland of large cities.