Next Article in Journal
Spatial–Temporal Evolution Characteristics of Landscape Ecological Risk in the Agro-Pastoral Region in Western China: A Case Study of Ningxia Hui Autonomous Region
Previous Article in Journal
Determinants of Farmers’ Behavior towards Land Conservation Practices in the Upper Citarum Watershed in West Java, Indonesia
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Land
Volume 11
Issue 10
10.3390/land11101828
land-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Looking Closer at the Patterns of Land Cover in the City of Porto, Portugal, between 1947 and 2019—A Contribution for the Integration of Ecological Data in Spatial Planning

by
Filipa Guilherme
1,2,*,
Eva García Moreno
2,
José Alberto Gonçalves
1,3,
Miguel A. Carretero
2,4 and
Paulo Farinha-Marques
1,2
1
Department of Geosciences, Environment and Spatial Plannings, Faculty of Sciences, University of Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal
2
CIBIO Research Centre in Biodiversity and Genetic Resources, InBIO Associate Laboratory, BIOPOLIS Program in Genomics, Biodiversity and Land Planning, Campus Agrário de Vairão, Rua Padre Armando Quintas 7, 4485-661 Vairão, Portugal
3
CIIMAR—Interdisciplinary Centre of Marine and Environmental Research, Terminal de Cruzeiros de Leixões, Avenida General Norton de Matos s/n, 4450-208 Matosinhos, Portugal
4
Department of Biology, Faculty of Sciences, University of Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal
*
Author to whom correspondence should be addressed.
Land 2022, 11(10), 1828; https://doi.org/10.3390/land11101828
Submission received: 22 September 2022 / Revised: 9 October 2022 / Accepted: 14 October 2022 / Published: 18 October 2022
(This article belongs to the Section Urban Contexts and Urban-Rural Interactions)
Browse Figures
Versions Notes

Abstract

:
As more people reside in cities and metropolitan areas, urban vegetation assumes an increasingly important role as one the main providers of ecosystem services in close proximity to human agglomerations. To improve the conditions for citizens and to optimise the sustainability of urban areas, the fields of landscape and urban ecology need to address the urgent priority to integrate ecological data in spatial planning, design, and management programs. With the objective to produce “actionable knowledge” for urban planning in the city of Porto (Portugal), we analyse the evolution of land cover since the mid-20th century at a fine spatial scale. Porto has followed the global trends of urbanisation, marked by a general increase in built-up and impervious surfaces that conquered the previously rural surrounding areas. This caused a severe decline in vegetation cover (especially herbaceous), as well as an increase in fragmentation and isolation of the remaining vegetation patches. These outcomes provide a detailed analysis of the city spatial dynamics, generating valuable information that can be relevant for future interventions regarding urban landscape change at a local scale, the most relevant for planning.

1. Introduction

The concept of landscape is complex and manifold, having been appropriated by numerous disciplines, from the arts and humanities to the natural sciences [1,2,3]. For the purpose of this study, landscape is “an area, as perceived by people, whose character is the result of the action and interaction of natural and/or human factors” [4]. The landscape is inherently complex, heterogenous, and dynamic [1,5,6] and it requires a strong simplification into smaller spatial units to be comprehensively analysed [3,7]. The landscape is characterised not only by the physical attributes of the land but also the human experience of it, i.e., by its composition, pattern, and relationship between its components, as well as the perception of all these factors [6,8].
Urbanisation is one of the main generators of change in landscape patterns and composition, as it generally implies the spread and/or densification of built-up and impervious surfaces and the loss and fragmentation of vegetation, natural habitats, and agricultural land [9,10]. This, in turn, negatively affects the biodiversity levels [11,12,13], environmental conditions [14,15,16,17,18,19] and quality of life in urban areas [20,21,22]. Particularly in urban environments, vegetation can be considered one of the most important contributors to environmental conditions, impacting the regulation of temperature, mitigation of flooding episodes, carbon sequestration and storage, adsorption of pollutants, provision of habitats for biodiversity, and supply of green spaces for recreation [14,20,23,24].
Cities and metropolitan areas are becoming more and more important in the context of the evolution of the human global population. In the mid-20th century, urban areas were populated by only around 30% of the global population (0.75 billion people); nowadays, around 55% of all humans live in urban settlements (4.22 billion people), as reported by the United Nations [25]. In 2050, it is estimated that around 68% of the world’s people will be urban dwellers (6.68 billion)—in fact, it is expected that all of the future growth in world population will be located in urban areas [25].
In the face of the rise in urban population, the Sustainable Development Goals of the United Nations highlight the need to pursue sustainability in urban centres, through the creation or transformation of cities so that they are greener, less polluted, and more resilient to climate change, ecological invasions, and emergent diseases, in order to increase the health and quality of life of their citizens and minimise the effects of natural disasters [26,27,28]. To achieve sustainable cities, the discipline of Landscape Ecology can play a central role in the multidirectional transfer of knowledge between ecological research and urban planning [29]. However, when compared to their natural counterparts, urban ecosystems generally present a very complex, heterogeneous, and dynamic landscape, driven by a multitude of interacting social, economic, and environmental factors, which translates into additional challenges for spatial planning and management [30]. Furthermore, there is often a mismatch of temporal and spatial scales, where the time and land units for planning and management decisions do not coincide with the ecological processes [29,30]. These factors can introduce barriers to the “knowledge to action” process, where both ecological research and policies at a broad scale (international or national level) are difficult to translate to the local level [29,31,32]. The “local level” can be defined as the physical areas that are directly intervened by urban planners, designers, managers, and users [29]. In urban contexts, the ecological dynamics of the landscape are largely affected by local scale decisions such as the planning and design of individual sites, infrastructures, and community masterplans [33]. At the local level, historical data about the environmental conditions of each site can also present important ecological opportunities and constraints; for example, historic ecological data often serve as an inspiration for urban design and landscape architecture projects, regarding ecosystem performance and aesthetic expression [33].
Most ecological research is focused on the advancement of academic knowledge, intended to inform policies at a broad scale, but with limited value for landscape planning at the city or neighbourhood level. Accordingly, most spatial planning and management programs fail to integrate ecological data [32,33,34]. Thus, it is imperative that scientists, planners, designers, communities, and other local stakeholders work collaboratively, creating and transferring knowledge in a format that is clear and useful for every thematic field, in an effort to advance the field of sustainable landscape planning [31,32,35,36,37].
In highly heterogeneous and dynamic regions, such as urban areas, ecological research must adapt a fine spatial scale and seek detailed temporal information that can be complemented with diverse social, economic, cultural, and political data, which are crucial to inform landscape planning and design, adjusted to the local context of each city [29,33]. In this sense, the study of land cover facilitates the quantitative analysis of the spatial patterns of vegetation and acts as a proxy for habitat distribution [20]. Land cover can be defined as the physical material that covers the surface of the landscape, usually related to vegetation types or artificial structures [38]. The concept of land cover is easily applicable everywhere as it transcends local conditions [39], but on the other hand, it is relevant at a local scale as it can be directly transformed through most urban interventions, not necessarily requiring a change in land use type. In contrast, land use classification, often used in urban ecological studies, is difficult to associate with biodiversity, habitat, and environmental conditions, as the relationship with vegetation cover remains unclear.
These challenges seem to be particularly relevant in compact cities [40,41,42,43]. As these undergo processes of densification, adequate policies regarding sustainable urban planning, advocating for habitat and biodiversity conservation while offering a suitable quality of life for residents, become imperative as there is (literally) no room for error. In this urban context, there is a pressing need for a concerted effort between multiple disciplines and professionals to generate useful knowledge, closer to the scale of intervention, in order to come up with innovative and tailor-made actions fitted to the local context.
With this in mind and using a compact city from the temperate region as a case study (Porto, Portugal), we analysed the change in urban land cover for an approximate period of 70 years. Our general intention was to produce “actionable knowledge” [35] with a fine spatial scale and vast temporal span. Thus, our research objectives were to (1) assess the land cover of Porto for three representative dates encompassing the main landscape changes, from the mid-20th century to recent days; (2) analyse the evolution of landscape pattern and composition with spatial metrics; and (3) identify the main trends, at a small scale, that could inform local plans and support direct actions taken by local governance and management systems for a more sustainable city.

2. Materials and Methods

2.1. Study Area

This study focuses on the Municipality of Porto, Portugal, in Southwestern Europe (Figure 1). The city is located in Northern Portugal, at the intersection of the Douro River with the Atlantic Ocean, where the temperate climate, with mild temperatures year-round and higher precipitation in the winter (Csb regime, according to the Koppen–Geiger classification system [44]), benefits a diverse community of plant species and habitats for European standards. The latest climate records indicate a mean annual temperature of 15.3 °C and total annual precipitation of 1174.0 mm (data for 2021 [45]).
The region has been occupied since pre-historic times, but it remained a small and compact urban centre, surrounded by agriculture and forest, until the 19th century. In contrast, the industrial development in the 20th century boosted the expansion of the built-up area to its administrative limits, forming an urban continuum with the adjoining municipalities [46]. Despite its limited size (around 42 km2 and almost 232,000 inhabitants [47]), Porto is currently the dynamic core of a metropolitan area composed of 17 municipalities, housing 1.7 million residents [48].

2.2. Land Cover Mapping and Classification

One of the main purposes of this study was to quantify the change in land cover of Porto considering a substantial time interval, based on historic remote sensing data. With this in mind, we decided to use three moments in time: one for the present time, one for a distant past (going as far back as possible with reliable data), and one intermediate moment. As far as our knowledge goes, the oldest aerial photography covering the city of Porto dates back to 1939, and it is available for consultation in the archives of the municipality. However, that first survey had substantial gaps, not covering the entire territory, and was executed during the winter, which makes it difficult to record deciduous vegetation; for these reasons, the aerial photograph of 1939 was not used for this study. The next oldest record corresponds to the 1947 photography obtained from Centro de Informação Geoespacial do Exército. The recent records of land cover in Porto were analysed based on the 2019 satellite photography, accessed through Google Earth. For the intermediate moment, considering similar intervals from the first and to the last moment, the 1979 aerial photography was selected, and it was made available for this study through a collaboration with the Municipality of Porto. Thus, the land cover for the city of Porto was analysed, through the interpretation of aerial and satellite imagery, in vectorial format in GIS software (ArcMap Desktop 10.5) for three different years: 1947, 1979, and 2019.
In order to have aerial photos as georeferenced raster layers in a GIS, original images must be orthorectified [49,50]. In the case of historical photos, this task may pose several difficulties due to the lack of knowledge about technical aspects of the aerial cameras used and also due to the difficulty in finding common points to present time images for a proper image georeferencing. Pinto et al. [50] propose a methodology based on individual photos, requiring a minimum number of 6 ground control points (GCPs) per photo. In the case of the 1947 photos, in a total of 24, finding the required number of GCPs would be difficult due to the changes that occurred in the city. Another methodology, proposed by Gonçalves [49], applicable to a standard coverage of aerial photographs with 60% overlap, requires a minimum of 4 points to georeference a block of images. It makes use of the structure from motion (SfM) strategy, now largely used for orthorectifying images acquired by drones. The method was applied with the Agisoft Metashape program [49]. It starts with the application of computer vision algorithms to find common points between overlapping images, generates a point cloud and a 3D model of the terrain, which is georeferenced in a block by a set of GCPs that perform 3D translation, rotation, and scale. Figure 2a,b represents a typical GCP in a garden that kept its original shape in the 1947 photo and in present time orthoimages. A set of 6 GCPs were used, resulting in residuals which can be used as a quality control of the georeferencing process. A root mean square error smaller than 2 m was obtained, which was considered appropriate for the analysis to be carried out. This is followed by a set of additional processing steps, resulting in a final orthomosaic, which can be inserted in the GIS and overlaid with other georeferenced data. Figure 2c shows a sample of the orthomosaic with vector polygons of building blocks. The quality of the orthomosaic geolocation was considered appropriate for a precise combination with other GIS data layers.
Land cover was classified with a version of the UrHBA method [39], streamlined for exclusive use with remote sensing data, especially considering the low resolution and black/white colour of the older photographs. UrHBA is a spatially explicit method, based on the visual interpretation of the landscape, designed to describe urban environments in a fine scale, making it particularly oriented for scientific research, urban planning, and data communication. Unlike traditional classifications of urban land that rely on land use data, UrHBA is based exclusively on land cover, which has a direct relationship with the presence or absence of vegetation. The classification of land units based on life traits of vegetation (instead of species or botanical communities) makes the method easily applicable in a wide range of geographical contexts. If necessary, the method allows for a more thorough characterisation later on, with the addition of detailed attributes, which should be obtained through fieldwork. The land cover categories used in this study are (1) Artificial Built Elements, ABE; (2) Trees and Shrubs, TRS; (3) Herbaceous, HER; (4) Sparsely Vegetated—Terrestrial, SPV; and (5) Sparsely Vegetated—Aquatic, AQU (Table 1). These were adapted from Urban Habitat Categories and super-categories from UrBHA (for further detail and definition of the categories, see [39,51]). The mapping procedure was also simplified—all elements were mapped as patches, even if with a linear shape; the minimum dimension for patches to be distinguished was 2000 m2, and the minimum width was 5 m.

2.3. Spatial Metrics

Land cover patterns were monitored by means of spatial metrics, which can be used as indicators of habitat availability, shape, fragmentation, and connectivity. For the purpose of the generation of spatial metrics, the term “patch” corresponds to the smaller spatial unit with homogeneous land cover, “class” consists of the group of all patches with the same land cover, and “landscape” identifies the entire study area, formed by a heterogeneous mosaic of patches [52].
Spatial metrics were calculated with V-LATE 2.0 beta [53] for the three resulting maps of land cover classification, at landscape, class, and patch levels. For landscape-level analysis, all patches are considered for the calculations, i.e., all the different land cover categories are included in the same analysis; at the class-level analysis, individual calculations are performed for each land cover category; and at patch-level, a series of metrics is generated for each land cover patch. This software uses vectorial data for the calculation of the most common metrics, and a few were selected as indicators of composition and configuration of the urban landscape, as shown in Table 2. These metrics can be grouped as (1) area and edge metrics—fundamentally related to the basic dimensions of patches, such as area and perimeter, and important for the analysis of habitat availability, fragmentation, and edge effects; (2) shape metrics—related to patch morphology and geometric complexity, which can influence edge effects, animal migration, and plant colonisation processes; (3) aggregation metrics—these indicate the tendency of patch types to be more or less dispersed and separated in the landscape, also known as “landscape texture”, influencing a diverse array of ecological processes, especially those concerning habitat loss, fragmentation, and isolation; (4) diversity metrics—important to understand the landscape composition, and only calculated for the landscape level of analysis [52].
Finally, a Kruskal–Wallis test was performed on SPSS Statistics v.27 software to check for significant differences between patch metrics for the three moments. The Kruskal–Wallis test is a nonparametric method for testing if multiple groups follow the same distribution. Next, a post hoc Dunn’s test was used for pairwise comparisons, in the case of rejection in the Kruskal–Wallis test, to identify which date shows a significant difference. These results were then interpreted considering the landscape as a whole and for each individual land cover class and thus are presented together with the landscape- and class-level analyses.

3. Results

3.1. Land Cover

Between 1947 and 2019, the city of Porto experienced an overall trend of an increase in artificial areas, at the expense of vegetated areas (Figure 3 and Figure 4). In 1947, the municipality of Porto exhibited two key urban areas, represented here as agglomerations of the land cover class of Artificial Built Elements (ABE): a smaller one, in the southwest, near the estuary of Douro River, and the main urban core, in a more central-southern position. These built-up areas have greatly expanded throughout the years, resulting in the duplication of the urban surface (from 31% in 1947 to 62% in 2019), currently forming an almost continuous mesh. The stability in the number of patches of ABE, along with the substantial increase in the proportion of the landscape (Figure 3), indicates that no new separate urban areas were created, but rather the existing ones were enlarged.
Large patches of arboreal vegetation exist mainly in the peripheral areas of Porto, especially in the western and eastern extremes, with most of them being preserved since the mid-20th century (Figure 4). This caused the percentage of woody vegetation (TRS) within the city, as well as the number of patches, to remain virtually unchanged from 1947 to 2019 (between 22% and 25% cover, and between 750 and 783 patches; Figure 3). On the other hand, the city has experienced a drastic change in the proportion of herbaceous cover (HER): in 1947, the landscape was clearly dominated by herbaceous vegetation, with approximately 40% cover, which reduced to 10% in 2019 (Figure 3 and Figure 4). Through visual analysis of the land cover maps, but mainly by means of the interpretation of the increase in the number of patches, accompanied by the reduction in the percentage cover, it becomes clear that herbaceous habitats were the most affected in terms of availability and connectivity.
Although the classes of Sparsely Vegetated—Terrestrial (SPV) and Sparsely Vegetated—Aquatic (AQU) are considered as components of ecologically significant habitats, they have remained consistent through the time period 1947–2019 and are residual in this study area (always less than 5%; Figure 3). Thus, SPV and AQU will not be analysed in detail as they play a reduced role in the interpretation of land cover evolution in the city of Porto.

3.2. Spatial Metrics

3.2.1. Landscape-Level Analysis

At the landscape level, the overall perspective of spatial metrics seems to point to a general trend of increased complexity and fragmentation (Figure 5; see also Table A1). Although the analysed study area remains the same for every year, the number of patches (NP) has increased over time, resulting in the decline of mean patch size (MPS). The increase in mean patch edge (MPE) and edge density (ED), as well as in all shape metrics (mean perimeter–area ratio: MPAR, mean shape index: MSI, and mean fractal dimension: MFD), reveal a general increase in patch shape irregularity and intricacy.
The values of mean Euclidean nearest neighbour distance (MENND) have increased throughout the analysed period, pointing to increased fragmentation and isolation of patches. Conversely, the values of mean proximity index (MPROX) have largely increased, suggesting higher proximity to large patches of the same land cover class. These results should be examined in more detail for a clearer interpretation (see class-level results below).
The territory seems to have experienced a decline in the diversity of land cover classes, especially during the transition from 1979 to 2019, as exposed by the decreasing values of the Shannon diversity index (SHDI) and the Shannon evenness index (SHEI). This indicates that one or more classes have become increasingly dominant over other classes, which can be better understood in the class-level analysis below.
Considering individual patch metrics, analysed for the entire landscape, the Kruskal–Wallis test revealed that most metrics showed significant differences for at least one of the periods; however, the mean values for Euclidean nearest neighbour distance did not differ significantly throughout the years (Table 3).

3.2.2. Class-Level Analysis

The spatial metrics for each individual class provide an overall image of increasing dominance of built-up areas (Figure 6; see also Table A2, Table A3 and Table A4). As previously explained, the results for SPV and AQU are not presented nor discussed, as their relevance in the land cover dynamics of the territory is negligible at this scale of analysis.
Artificial Built Elements (ABE) cover. As expected, following the global urbanisation trends of the 20th century, the class of ABE showed a substantial increment in area, as was revealed by the area metrics: class area (CA), class proportion (CP), largest patch index (LPI), mean patch size (MPS), and mean patch edge (MPE) gradually increase over time. Accordingly, the results in terms of aggregation metrics show a trend of decreasing MENND and increasing MPROX, which suggests higher connectivity among ABE patches over time. Shape complexity remained relatively constant throughout the years, as revealed by the evolution of shape metrics: MPAR, MSI, and MFD show similar values for the three time periods (Figure 6, Table A2, Table A3 and Table A4). Regarding the analysis of individual patch metrics for ABE land cover, the Kruskal–Wallis test revealed that most spatial metrics do not significantly differ in the three different moments, with the exception of Euclidean nearest neighbour distance and proximity index (Table 4).
Herbaceous (HER) cover. Area metrics (CA, CP, LPI, MPS, and MPE) reveal a severe decrease in the availability of HER land cover, while the increasing values of ED, as well as MPAR, MSI, and MFD, seem to point to a situation of increased patch shape complexity and irregularity over time. The evolution of aggregation metrics also suggests a rise in isolation and fragmentation of HER patches, as can be especially seen in the results of increasing values of MENND and decreasing values of MPROX (Figure 6, Table A2, Table A3 and Table A4). According to the Kruskal–Wallis test, all patch metrics for HER land cover class show significant differences between the analysed time periods (Table 5).
Trees and Shrubs (TRS) cover. According to most area metrics (CA, CP, LPI, MPS), the land cover dominated by TRS apparently maintained a fairly constant area throughout the years. The evolution of aggregation metrics (NP, MENND, MPROX, and division index, DIV) also suggests the higher stability of this land cover type regarding availability and connectivity. However, the increasing values of edge (MPE, ED) and shape metrics (MPAR, MSI, MFD) reveal a scenario of increased shape complexity and irregularity (Figure 6, Table A2, Table A3 and Table A4). Indeed, the Kruskal–Wallis test revealed that patch size and edge have significantly different values for the three analysed years, in addition to patch shape metrics (perimeter–area ratio, shape index, and fractal dimension). On the other hand, aggregation metrics for TRS have not changed significantly throughout the years (Table 6).
General overview. When analysing the city at a broad scale, there is an obvious trend of urban sprawl and densification, but this was translated with different spatial patterns in distinct portions of the city. Four main tendencies of land cover evolution were identified throughout the city, which are represented in Figure 7: (i) dominance of ABE from at least 1947 to the present, mostly visible in the city centre; (ii) drastic urban expansion, especially from 1979 onward, with the consequent elimination of HER cover, particularly noticeable in the northern section; (iii) slight increase in ABE surface, but with consequences in terms of increased fragmentation and irregularity of TRS and HER cover, as can be seen, for example, in the western section; and (iv) decrease in HER cover associated with increase in TRS cover, also coupled with moderate increase in ABE, as occurred in the eastern area.

4. Discussion

The city of Porto epitomises many of the urbanisation trends in vegetation cover observed in many European cities, having experienced a notorious sprawl and densification of its urban fabric in the past few decades [9,54,55]. In the first half of the 20th century, its territory encompassed a small city centre, as well as a few ramifications to even smaller urbanised clusters in the periphery, and was still being dominated by vegetation cover, especially herbaceous vegetation in agricultural fields. The second half of the 20th century saw an acceleration of urbanisation, and at the beginning of the 21st century, the urban fabric occupied most of the territory contained by the boundaries of Porto. The former rural areas that surrounded the historical urban centres were progressively conquered by new urbanisation processes, which caused the increased dominance of buildings and sealed surfaces (ABE cover), as well as the reduction in diversity of land cover types. Most of the herbaceous patches (HER cover) have disappeared or have been extremely fragmented and isolated from similar patches; this is mainly due to the reclamation of land for new urbanisation, but also related to the abandonment of agricultural practices. Nowadays, most of the remaining vegetation cover persists in patches of trees (TRS cover), which reveals some concern with the preservation of tree cover; in fact, while some new tree cover originated from the ecological succession in abandoned or unmanaged lots, most of the larger tree-covered patches that existed in the mid-20th century were integrated in new public parks.
These results are in line with previous studies performed for the Porto region, although for different contexts, time frames, and scales of approach. For example, in their analysis of the evolution of green spaces in Porto in the 20th century, Madureira et al. [56] demonstrated a severe decrease in agricultural areas, accompanied by a general increase in green space destruction and fragmentation. Similarly, Bagan and Yamagata [57] revealed an association between the urban growth of Porto and the reduction in several types of vegetation cover, within a shorter, more recent period, from 1984 to 2009. Several other cities, in Europe and worldwide, also present an urbanisation scenario similar to that experienced by Porto in the past few decades: an increase in built-up and impervious areas and a decline in vegetation and green space cover (e.g., [9,58,59,60]); however, it should be noted that most of these studies consider only short-term changes, rarely encompassing more than 20 or 30 years of urban land cover evolution. With a similar time frame to our study, Lennert et al. [61] explored the past dynamics of the Budapest metropolitan region from 1959 to 2012 and obtained similar results: the growth in artificial surfaces caused the decline in vegetated areas, such as natural or agricultural land; nevertheless, the large-scale approach can make it difficult to translate to local actions in each of the individual municipalities. It is worth mentioning that a lot of cities, especially large European and North American cities, have been able to counteract this tendency of vegetation cover loss since the late 20th century [9,62]. While some of these trends can be explained by the rise in temperatures favouring vegetation growth, many of these cities have in fact strengthened their green infrastructure and promoted the application of nature-based solutions [63]. This could also be the case of Porto if the outcomes of this study are properly integrated in urban planning.
The fine scale of analysis allows for a detailed look at specific portions of the city, which can be particularly relevant to understand the context and implications of future interventions in urban areas, which may be of significant interest for urban planners and decision makers. For example, the city centre (Figure 7a) has been dominated by ABE for a long time, with a fairly constant area, and TRS cover also seems to be quite persistent. This may be suggestive of some habitat and green space conservation efforts, as most of the TRS patches that were already present in 1947 belonged to public parks, gardens, or green squares that were maintained to the present day; some private green spaces inside blocks of buildings have also been preserved. Thus, it is essential that, in this densely built area, all pervious land is protected from future construction and new opportunities for the increment of vegetation cover are seized and planned carefully to maintain appropriate levels of vegetation diversity.
On the other hand, the northern part of Porto (Figure 7b) reveals an obvious scenario of rapid urban expansion where agricultural fields, dominated by HER cover, were progressively displaced by urbanisation, seemingly without much planning or conservation concerns. In this section of the city, the loss and fragmentation of vegetated areas become quite noticeable, especially due to large continuous patches of HER cover, and also TRS cover to a lesser extent, being encroached by the development of new buildings, roads, and other infrastructures (ABE). It is recommended that future interventions in this or other areas with a similar profile take actions to preserve, enlarge, and connect the remnant patches of vegetation, in order to enhance the ecological performance of the urban ecosystem.
The western extreme of the city (Figure 7c) also shows a significant decrease in the proportion of HER cover due to the decline in agricultural activities. The scenario of increase in fragmentation and shape irregularity of TRS cover is particularly visible in this portion of the city, as large patches of tree-covered private gardens and backyards, present in 1947, were gradually replaced by a denser matrix of buildings; most of the TRS patches in 2019 are tree-lined streets and narrow fragments of private gardens. In this area of the city, there are still some relatively large patches dominated by herbaceous vegetation that represent some of the few cultivated areas in the whole city that persist today. Considering the future urban developments expected for this area, we recommend that this portion of land retains the highest degree of perviousness as possible, protecting one of the few areas of the city with less disturbed soil. In this regard, we highlight two major options: (1) the preservation of agricultural spaces, for example, in the form of community gardens or allotments, and (2) the establishment of herbaceous habitats, based mainly on native plant species. These practices could contribute to the conservation of soil and water quality, while also promoting the biodiversity of grassland and farmland habitats, which is declining at the European level [64].
The eastern portion of Porto (Figure 7d) was also dominated by HER cover in 1947, mainly in the form of crops and pastures. Although some new urban developments occurred, namely the construction of a few residential neighbourhoods and large roads, this portion remains one of the least urbanised of the city. However, the area of HER cover has also decreased substantially, especially due to the abandonment of agricultural fields, which caused an increment in mostly spontaneous TRS cover. This can pose an additional challenge due to the proliferation of invasive species that may follow the abandonment of these fields. Nevertheless, the extent of vacant land present in this portion of the city can generate opportunities for the creation of new accessible green spaces that provide numerous benefits for the urban population, especially considering the lack of public green spaces amongst this intricate urban fabric with an accentuated rural character.
Due to their higher degree of complexity and heterogeneity, the smaller scale of approach seems to be particularly relevant in urban areas, where there is a strong competition for space among different interest groups and where decisions at a small scale can have significant ecological or social effects. For instance, decisions such as the preservation of even a small patch of vegetation can dictate the survival of a population of a certain species or can contribute to the experience of nature for the nearest residents, improving their health, well-being, and quality of life at a scale that even transcends that particular site.
To complement these results, further data could be generated encompassing, for example, a more detailed classification of vegetation types, a finer temporal scale (considering shorter time intervals), and the study of land cover trajectories. Additionally, depending on the aim of research, but particularly relevant for spatial planning, it may be worth analysing the relationship between land cover patterns and evolution with biodiversity, quality of life, health and well-being, and other factors.
As such, this research can be easily reproduced in a variety of urban areas, as the methods are quite simple and do not require prior specialised training for the visual interpretation of the landscape. However, due to the detailed scale of analysis, the method can become quite time-consuming when considering larger study areas; this could be overcome with the creation of automatic algorithms for the classification of land cover, which was outside of the scope of this research.

5. Conclusions

The city of Porto experienced evident processes of urban sprawl and densification in the last century, mimicking other cities worldwide. Thus, the findings and recommendations of this research can also be applied to a variety of cities with a similar context, particularly other European cities that are compact and undergoing phases of infill development. To counteract the loss of vegetation, especially grassland and farmland habitats, we propose that these cities invest in the conservation, restauration, or creation of herbaceous patches, in the form of meadows, allotments, urban farming, and other amenities, that could be encouraged in public and private areas. Arboreal vegetation is also important, but its conservation is a more common practice (that should be maintained). Cities that aim to be sustainable should consider several aspects regarding green spaces: maximise, to the greatest possible extent, the amount of vegetation cover in the city; enhance the diversity of green spaces and in green spaces (e.g., include different types of green spaces, with different uses and different compositions); optimise the relation of vegetation patches with each other and with the surrounding built environment; and consider the temporal variation or continuity of vegetation.
The small-scale approach of this research brings the results closer to the scale of intervention, i.e., it allows a focus, with great detail, on particular sites, contributing to decision making in future interventions in the urban landscape (e.g., preservation of the existing vegetation, selection of the best habitats to implement, etc.). This way, the scale of approach and the emphasis on vegetation facilitate the translation of scientific research into concrete practical measures, and not just general recommendations on a broad geographic scale that are difficult to implement at the local decision scale.
This study showed that it is possible to obtain urban land cover data at a fine spatial scale, while also considering the history of the vegetation on each location. This research was outlined with urban ecology and urban planning in mind, but it can also be useful to other fields of scientific research and professional practices, such as architecture, landscape architecture, geography, social studies, etc., that take interest in improving the urban environment and the health and quality of life for urban residents.

Author Contributions

Conceptualisation, F.G., M.A.C. and P.F.-M.; Formal analysis, F.G.; Funding acquisition, F.G., E.G.M., M.A.C. and P.F.-M.; Investigation, F.G. and E.G.M.; Methodology, F.G., M.A.C. and P.F.-M.; Resources, J.A.G.; Supervision, M.A.C. and P.F.-M.; Visualisation, F.G.; Writing—original draft, F.G.; Writing—review and editing, J.A.G., M.A.C. and P.F.-M. All authors have read and agreed to the published version of the manuscript.

Funding

F.G. was funded by Fundação para a Ciência e a Tecnologia (FCT, Portugal), through the PhD grant SFRH/BD/137208/2018, originating from MCTES, FSE and EU funds. E.G.M. was funded by the Chamber of Commerce of Alicante, through the Mobility Plan—Integral Qualification and Employment Programme (PICE).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Acknowledgments

We would like to thank the Municipality of Porto, namely Pedro Pombeiro from the Department of Environmental Planning and Management and Isabel Martins from the Department of Urban Planning, for their support of this research and for facilitating the connection with various municipal teams. We would also like to thank the Municipal Division of Geographic Information, particularly Alexandra Rodrigues and Célia Azevedo, for their collaboration on the data collection in the historical aerial photograph archives.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Appendix A. Values of Spatial Metrics

Table
Table A1. Spatial metrics at the landscape level, for 1947, 1979, and 2019.
Table A1. Spatial metrics at the landscape level, for 1947, 1979, and 2019.
TypeMetric194719792019
Area and edge metricsTA: Total area (ha)4160.754160.754160.75
LPI: Largest patch index (%)29.0243.3360.45
MPS: Mean patch size (ha)3.112.742.68
MPE: Mean patch edge (m)1477.241637.281646.62
ED: Edge density (m/ha)474.33598.13615.39
Shape metricsMPAR: Mean perimeter–area ratio 0.1020.1210.141
MSI: Mean shape index2.2352.6042.844
MFD: Mean fractal dimension1.4511.4891.518
Aggregation metricsNP: Number of patches133615201555
MENND: Mean Euclidean nearest neighbour distance (m)25.4029.8432.78
MPROX: Mean proximity index9948.1842,631.88107,968.34
Diversity metricsCR: Class richness555
SHDI: Shannon’s diversity index1.2701.2451.043
SHEI: Shannon’s evenness index0.7890.7730.648
Table
Table A2. Area and edge metrics at the class level, for 1947, 1979, and 2019.
Table A2. Area and edge metrics at the class level, for 1947, 1979, and 2019.
MetricClass194719792019
CA: Total class area (ha)ABE1280.831888.412597.82
TRS988.491034.21923.29
HER1653.231004.87427.03
SPV71.8476.5054.89
AQU166.35156.75157.71
CP: Class proportion (%)ABE30.7845.3962.44
TRS23.7624.8622.19
HER39.7324.1510.26
SPV1.731.841.32
AQU4.003.773.79
LPI: Largest patch index (%)ABE29.0243.3360.45
TRS0.661.030.93
HER2.821.430.22
SPV0.450.490.92
AQU3.162.222.10
MPS: Mean patch size (ha)ABE10.4115.7421.65
TRS1.321.311.18
HER4.081.820.70
SPV1.361.421.83
AQU33.2731.359.86
MPE: Mean patch edge (m)ABE6015.938559.079474.38
TRS863.691111.261223.57
HER1264.95923.00665.25
SPV767.90801.37830.30
AQU6570.876441.962341.89
ED: Edge density (m/ha)ABE577.72543.89437.65
TRS655.31846.711037.65
HER309.88507.94944.06
SPV566.52565.67453.80
AQU197.50205.49237.59
Table
Table A3. Shape metrics at the class level, for 1947, 1979, and 2019.
Table A3. Shape metrics at the class level, for 1947, 1979, and 2019.
MetricClass194719792019
MPAR: Mean perimeter–area ratioABE0.1850.1530.168
TRS0.1010.1300.153
HER0.0790.1020.125
SPV0.0960.1090.098
AQU0.1260.0460.066
MSI: Mean shape indexABE2.8172.9652.909
TRS2.2632.8293.259
HER2.0102.2352.361
SPV2.0552.2221.915
AQU3.7603.2472.071
MFD: Mean fractal dimensionABE1.5181.5121.521
TRS1.4611.5131.546
HER1.4141.4531.489
SPV1.4391.4591.436
AQU1.5011.4161.389
Table
Table A4. Aggregation metrics at the class level, for 1947, 1979, and 2019.
Table A4. Aggregation metrics at the class level, for 1947, 1979, and 2019.
MetricClass194719792019
NP: Number of patchesABE123120120
TRS750788783
HER405553606
SPV535430
AQU5516
MENND: Mean Euclidean nearest neighbour distance (m)ABE29.5120.736.71
TRS18.7317.3617.63
HER21.3027.4730.72
SPV126.96247.82408.95
AQU181.13123.33342.81
MPROX: Mean proximity index (for a 100 m buffer)ABE87,302.27526,604.251,389,437.08
TRS647.791340.991299.65
HER5036.90964.48201.06
SPV161.3169.81187.65
AQU3650.342822.40826.43
DIV: Division indexABE11.108.856.24
TRS99.3799.1999.23
HER97.9598.6499.59
SPV89.0188.2650.54
AQU35.1359.0564.32

References

  1. Forman, R.T.T.; Godron, M. Patches and Structural Components For A Landscape Ecology. Bioscience 1981, 31, 733–740. [Google Scholar]
  2. Doherty, G.; Waldheim, C. What Is Landscape? In Is Landscape… ? Essays on the Identity of Landscape; Routledge: London, UK, 2016; ISBN 978-1-138-01847-1. [Google Scholar]
  3. Simensen, T.; Halvorsen, R.; Erikstad, L. Methods for Landscape Characterisation and Mapping: A Systematic Review. Land Use Policy 2018, 75, 557–569. [Google Scholar] [CrossRef]
  4. Council of Europe Landscape Convention. Eur. Treaty Ser. 2000, 176, 1–8.
  5. Risser, P.G. Landscape Ecology: State of the Art. In Landscape Heterogeneity and Disturbance; Turner, M.G., Ed.; Springer: New York, NY, USA, 1987; ISBN 978-1-4612-9137-4. [Google Scholar]
  6. Naveh, Z. Ecosystem and Landscapes—A Critical Comparative Appraisal. J. Landsc. Ecol. 2012, 3, 64–81. [Google Scholar] [CrossRef] [Green Version]
  7. Bunce, R.G.H.; Barr, C.J.; Clarke, R.T.; Howard, D.C.; Lane, A.M.J. Land Classification for Strategic Ecological Survey. J. Environ. Manag. 1996, 47, 37–60. [Google Scholar] [CrossRef]
  8. Kaplan, R. The Analysis of Perception via Preference: A Strategy for Studying How the Environment Is Experienced. Landsc. Plan. 1985, 12, 161–176. [Google Scholar] [CrossRef] [Green Version]
  9. Richards, D.R.; Belcher, R.N. Global Changes in Urban Vegetation Cover. Remote Sens. 2020, 12, 23. [Google Scholar] [CrossRef] [Green Version]
  10. Güneralp, B.; Reba, M.; Hales, B.U.; Wentz, E.A.; Seto, K.C. Trends in Urban Land Expansion, Density, and Land Transitions from 1970 to 2010: A Global Synthesis. Environ. Res. Lett. 2020, 15, 044015. [Google Scholar] [CrossRef]
  11. Beninde, J.; Veith, M.; Hochkirch, A. Biodiversity in Cities Needs Space: A Meta-Analysis of Factors Determining Intra-Urban Biodiversity Variation. Ecol. Lett. 2015, 18, 581–592. [Google Scholar] [CrossRef] [PubMed]
  12. Norton, B.A.; Evans, K.L.; Warren, P.H. Urban Biodiversity and Landscape Ecology: Patterns, Processes and Planning. Curr. Landsc. Ecol. Reports 2016, 1, 178–192. [Google Scholar] [CrossRef] [Green Version]
  13. Rastandeh, A.; Brown, D.; Pedersen Zari, M. Biodiversity Conservation in Urban Environments: A Review on the Importance of Spatial Patterning of Landscapes. In Proceedings of the Ecocity World Summit, Melbourne, Australia, 12–14 July 2017. [Google Scholar]
  14. De Carvalho, R.M.; Szlafsztein, C.F. Urban Vegetation Loss and Ecosystem Services: The Influence on Climate Regulation and Noise and Air Pollution. Environ. Pollut. 2019, 245, 844–852. [Google Scholar] [CrossRef]
  15. Gatto, E.; Buccolieri, R.; Aarrevaara, E.; Ippolito, F.; Emmanuel, R.; Perronace, L.; Santiago, J.L. Impact of Urban Vegetation on Outdoor Thermal Comfort: Comparison between a Mediterranean City (Lecce, Italy) and a Northern European City (Lahti, Finland). Forests 2020, 11, 228. [Google Scholar] [CrossRef]
  16. Grigoraș, G.; Urițescu, B. Land Use/Land Cover Changes Dynamics and Their Effects on Surface Urban Heat Island in Bucharest, Romania. Int. J. Appl. Earth Obs. Geoinf. 2019, 80, 115–126. [Google Scholar] [CrossRef]
  17. Kafy, A.A.; Al Rakib, A.; Fattah, M.A.; Rahaman, Z.A.; Sattar, G.S. Impact of Vegetation Cover Loss on Surface Temperature and Carbon Emission in a Fastest-Growing City, Cumilla, Bangladesh. Build. Environ. 2022, 208, 108573. [Google Scholar] [CrossRef]
  18. Liu, Y.; Peng, J.; Wang, Y. Efficiency of Landscape Metrics Characterizing Urban Land Surface Temperature. Landsc. Urban Plan. 2018, 180, 36–53. [Google Scholar] [CrossRef]
  19. Melaas, E.K.; Wang, J.A.; Miller, D.L.; Friedl, M.A. Interactions between Urban Vegetation and Surface Urban Heat Islands: A Case Study in the Boston Metropolitan Region. Environ. Res. Lett. 2016, 11, 054020. [Google Scholar] [CrossRef] [Green Version]
  20. Dobbs, C.; Nitschke, C.; Kendal, D. Assessing the Drivers Shaping Global Patterns of Urban Vegetation Landscape Structure. Sci. Total Environ. 2017, 592, 171–177. [Google Scholar] [CrossRef] [PubMed]
  21. Mears, M.; Brindley, P.; Jorgensen, A.; Ersoy, E.; Maheswaran, R. Greenspace Spatial Characteristics and Human Health in an Urban Environment: An Epidemiological Study Using Landscape Metrics in Sheffield, UK. Ecol. Indic. 2019, 106, 105464. [Google Scholar] [CrossRef]
  22. Wei, C.; Padgham, M.; Barona, P.C.; Blaschke, T. Scale-Free Relationships between Social and Landscape Factors in Urban Systems. Sustainability 2017, 9, 84. [Google Scholar] [CrossRef] [Green Version]
  23. Robinson, S.L.; Lundholm, J.T. Ecosystem Services Provided by Urban Spontaneous Vegetation. Urban Ecosyst. 2012, 15, 545–557. [Google Scholar] [CrossRef]
  24. Smith, A.C.; Harrison, P.A.; Pérez Soba, M.; Archaux, F.; Blicharska, M.; Egoh, B.N.; Erős, T.; Fabrega Domenech, N.; György, I.; Haines-Young, R.; et al. How Natural Capital Delivers Ecosystem Services: A Typology Derived from a Systematic Review. Ecosyst. Serv. 2017, 26, 111–126. [Google Scholar] [CrossRef] [Green Version]
  25. United Nations. World Urbanization Prospects. The 2018 Revision (ST/ESA/SER.A/420); United Nations: New York, NY, USA, 2019; ISBN 9789211483192. [Google Scholar]
  26. Gaertner, M.; Wilson, J.R.U.; Cadotte, M.W.; MacIvor, J.S.; Zenni, R.D.; Richardson, D.M. Non-Native Species in Urban Environments: Patterns, Processes, Impacts and Challenges. Biol. Invasions 2017, 19, 3461–3469. [Google Scholar] [CrossRef]
  27. Riley, M.F. One Health Pandemic Prevention and Mitigation: The Role of FDA. Food Drug Law J. 2021, 76, 200–234. [Google Scholar]
  28. United Nations. The Sustainable Development Goals Report; United Nations, Department of Economic and Social Affairs: New York, NY, USA, 2020. [Google Scholar]
  29. Opdam, P.; Nassauer, J.I.; Wang, Z.; Albert, C.; Bentrup, G.; Castella, J.C.; McAlpine, C.; Liu, J.; Sheppard, S.; Swaffield, S. Science for Action at the Local Landscape Scale. Landsc. Ecol. 2013, 28, 1439–1445. [Google Scholar] [CrossRef]
  30. Borgström, S.T.; Elmqvist, T.; Algelstam, P.; Alfsen-Norodom, C. Scale Mismatches in Management of Urban Landscapes. Ecol. Soc. 2006, 11, 16. [Google Scholar] [CrossRef] [Green Version]
  31. Childers, D.L.; Pickett, S.T.A.; Grove, J.M.; Ogden, L.; Whitmer, A. Advancing Urban Sustainability Theory and Action: Challenges and Opportunities. Landsc. Urban Plan. 2014, 125, 320–328. [Google Scholar] [CrossRef]
  32. Childers, D.L.; Cadenasso, M.L.; Grove, J.M.; Marshall, V. An Ecology for Cities: A Transformational Nexus of Design and Ecology to Advance Climate Change Resilience and Urban Sustainability. Sustainability 2015, 7, 3774–3791. [Google Scholar] [CrossRef] [Green Version]
  33. Brown, I.T. Managing Cities as Urban Ecosystems: Fundamentals and a Framework for Los Angeles, California. Cities Environ. 2017, 10, 4. [Google Scholar]
  34. Opdam, P. Using Ecosystem Services in Community- Based Landscape Planning: Science Is Not Ready to Deliver. In Landscape Ecology for Sustainable Environment and Culture; Fu, B., Jones, K.B., Eds.; Springer: Dordrecht, The Netherlands, 2013; pp. 77–101. [Google Scholar]
  35. Brunet, L.; Tuomisaari, J.; Lavorel, S.; Crouzat, E.; Bierry, A.; Peltola, T.; Arpin, I. Actionable Knowledge for Land Use Planning: Making Ecosystem Services Operational. Land Use Policy 2018, 72, 27–34. [Google Scholar] [CrossRef]
  36. Pickett, S.T.A.; Cadenasso, M.L.; Childers, D.L.; Mark, J.; Zhou, W.; Pickett, S.T.A.; Cadenasso, M.L.; Childers, D.L.; Mcdonnell, M.J.; Zhou, W. Evolution and Future of Urban Ecological Science: Ecology in, of, and for the City. Ecosyst. Heal. Sustain. 2016, 2, e01229. [Google Scholar] [CrossRef]
  37. Pauleit, S.; Hanse, R.; Rall, E.L.; Rolf, W.; Lierop, M. Van Green Infrastructure for the City of the Future. Perspectives from Europe. Work. Pap. Riv. Online Di Urban@It 2020, 2. [Google Scholar]
  38. Haines-Young, R. Land Use and Biodiversity Relationships. Land Use Policy 2009, 26S, S178–S186. [Google Scholar] [CrossRef]
  39. Farinha-Marques, P.; Fernandes, C.; Guilherme, F.; Lameiras, J.M.; Alves, P.; Bunce, R.G.H.H. Urban Habitats Biodiversity Assessment (UrHBA): A Standardized Procedure for Recording Biodiversity and Its Spatial Distribution in Urban Environments. Landsc. Ecol. 2017, 32, 1753–1770. [Google Scholar] [CrossRef]
  40. Haaland, C.; van den Bosch, C.K. Challenges and Strategies for Urban Green-Space Planning in Cities Undergoing Densification: A Review. Urban For. Urban Green. 2015, 14, 760–771. [Google Scholar] [CrossRef]
  41. Jim, C.Y.; Konijnendijk van den Bosch, C.; Chen, W.Y. Acute Challenges and Solutions for Urban Forestry in Compact and Densifying Cities. J. Urban Plan. Dev. 2018, 144, 04018025. [Google Scholar] [CrossRef]
  42. Bibri, S.E.; Krogstie, J.; Kärrholm, M. Compact City Planning and Development: Emerging Practices and Strategies for Achieving the Goals of Sustainability. Dev. Built Environ. 2020, 4, 100021. [Google Scholar] [CrossRef]
  43. Wolff, M.; Haase, D. Mediating Sustainability and Liveability—Turning Points of Green Space Supply in European Cities. Front. Environ. Sci. 2019, 7, 61. [Google Scholar] [CrossRef] [Green Version]
  44. Peel, M.C.; Finlayson, B.L.; McMahon, T.A. Updated World Map of the Köppen-Geiger Climate Classification. Hydrol. Earth Syst. Sci. 2007, 11, 1633–1644. [Google Scholar] [CrossRef] [Green Version]
  45. PORDATA PORDATA—Poluição Atmosférica e Clima. Available online: https://www.pordata.pt/subtema/portugal/poluicao+atmosferica+e+clima-86 (accessed on 30 September 2022).
  46. Ramos, L.A.d.O. História Do Porto; Porto Editora: Porto, Portugal, 2000; ISBN 2111100125628. [Google Scholar]
  47. INE Plataforma de Divulgação Dos Censos 2021—Resultados Provisórios. Available online: https://www.ine.pt/scripts/db_censos_2021.html (accessed on 21 March 2022).
  48. AMP AMPorto. Available online: http://portal.amp.pt/pt/ (accessed on 21 March 2022).
  49. Gonçalves, J.A. Automatic Orientation and Mosaicking of Archived Aerial Photography Using Structure from Motion. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci.-ISPRS Arch. 2016, 40, 123–126. [Google Scholar] [CrossRef] [Green Version]
  50. Pinto, A.T.; Gonçalves, J.A.; Beja, P.; Honrado, J.P. From Archived Historical Aerial Imagery to Informative Orthophotos: A Framework for Retrieving the Past in Long-Term Socioecological Research. Remote Sens. 2019, 11, 10–12. [Google Scholar] [CrossRef] [Green Version]
  51. Farinha-Marques, P.; Fernandes, C.; Guilherme, F.; Lameiras, J.M.; Alves, P.; Bunce, R. Morphology & Biodiversity in the Urban Green Spaces of the City of Porto. Book II—Habitat Mapping and Characterization; CIBIO—Centro de Investigação em Biodiversidade e Recursos Genéticos: Porto, Portugal, 2015; ISBN 9789899873247. [Google Scholar]
  52. McGarigal, K. FRAGSTATS Help; University of Massachussets: Amherst, MA, USA, 2015. [Google Scholar]
  53. Lang, S.; Tiede, D. vLATE Extension Für ArcGIS—Vektorbasiertes Tool Zur Quantitativen Landschaftsstrukturanalyse [vLate Extension for ArcGIS - Vector Based Tool for Quantitative Landscape Analysis]. In Proceedings of the 18th ESRI European User Conference, Innsbruck, Austria, 8–10 October 2003. [Google Scholar]
  54. Kasanko, M.; Barredo, J.I.; Lavalle, C.; McCormick, N.; Demicheli, L.; Sagris, V.; Brezger, A. Are European Cities Becoming Dispersed? A Comparative Analysis of 15 European Urban Areas. Landsc. Urban Plan. 2006, 77, 111–130. [Google Scholar] [CrossRef]
  55. Pauleit, S.; Ennos, R.; Golding, Y. Modeling the Environmental Impacts of Urban Land Use and Land Cover Change—A Study in Merseyside, UK. Landsc. Urban Plan. 2005, 71, 295–310. [Google Scholar] [CrossRef]
  56. Madureira, H.; Andresen, T.; Monteiro, A. Green Structure and Planning Evolution in Porto. Urban For. Urban Green. 2011, 10, 141–149. [Google Scholar] [CrossRef]
  57. Bagan, H.; Yamagata, Y. Land-Cover Change Analysis in 50 Global Cities by Using a Combination of Landsat Data and Analysis of Grid Cells. Environ. Res. Lett. 2014, 9, 064015. [Google Scholar] [CrossRef]
  58. Gupta, R. The Pattern of Urban Land-Use Changes: A Case Study of the Indian Cities. Environ. Urban. ASIA 2014, 5, 83–104. [Google Scholar] [CrossRef]
  59. Kaspersen, P.S.; Ravn, N.H.; Arnbjerg-Nielsen, K.; Madsen, H.; Drews, M. Comparison of the Impacts of Urban Development and Climate Change on Exposing European Cities to Pluvial Flooding. Hydrol. Earth Syst. Sci. 2017, 21, 4131–4147. [Google Scholar] [CrossRef] [Green Version]
  60. Timalsina, B.; Mavoa, S.; Hahs, A.K. Dynamic Changes in Melbourne’s Urban Vegetation Cover—2001 to 2016. Land 2021, 10, 814. [Google Scholar] [CrossRef]
  61. Lennert, J.; Farkas, J.Z.; Kovács, A.D.; Molnár, A.; Módos, R.; Baka, D.; Kovács, Z. Measuring and Predicting Long-Term Land Cover Changes in the Functional Urban Area of Budapest. Sustainability 2020, 12, 3331. [Google Scholar] [CrossRef] [Green Version]
  62. Corbane, C.; Martino, P.; Panagiotis, P.; Aneta, F.J.; Michele, M.; Sergio, F.; Marcello, S.; Daniele, E.; Gustavo, N.; Thomas, K. The Grey-Green Divide: Multi-Temporal Analysis of Greenness across 10,000 Urban Centres Derived from the Global Human Settlement Layer (GHSL). Int. J. Digit. Earth 2020, 13, 101–118. [Google Scholar] [CrossRef]
  63. Zhang, W.; Randall, M.; Jensen, M.B.; Brandt, M.; Wang, Q.; Fensholt, R. Socio-Economic and Climatic Changes Lead to Contrasting Global Urban Vegetation Trends. Glob. Environ. Chang. 2021, 71, 102385. [Google Scholar] [CrossRef]
  64. European Environment Agency Abundance and Distribution of Selected Species in Europe. Available online: https://www.eea.europa.eu/ims/abundance-and-distribution-of-selected (accessed on 30 September 2022).
Land 11 01828 g001 550
Figure 1. Location of the city of Porto, Portugal.
Figure 1. Location of the city of Porto, Portugal.
Land 11 01828 g001
Land 11 01828 g002 550
Figure 2. (a) Examples of a GCP in the 1947 aerial photo; (b) the same point in present time orthoimages; (c) large portion of the final orthomosaic, with overlaid vector data of building blocks in a GIS, showing the accurate geolocation of the orthomosaic.
Figure 2. (a) Examples of a GCP in the 1947 aerial photo; (b) the same point in present time orthoimages; (c) large portion of the final orthomosaic, with overlaid vector data of building blocks in a GIS, showing the accurate geolocation of the orthomosaic.
Land 11 01828 g002
Land 11 01828 g003 550
Figure 3. (a) Land cover class proportions in 1947, 1979, and 2019; (b) number of patches per class in 1947, 1979, and 2019 (see also Table A1).
Figure 3. (a) Land cover class proportions in 1947, 1979, and 2019; (b) number of patches per class in 1947, 1979, and 2019 (see also Table A1).
Land 11 01828 g003
Land 11 01828 g004 550
Figure 4. Land cover maps of Porto in 1947, 1979, and 2019.
Figure 4. Land cover maps of Porto in 1947, 1979, and 2019.
Land 11 01828 g004
Land 11 01828 g005 550
Figure 5. Evolution of spatial metrics at the landscape level, for the period 1947–1979–2019 (see Table A1 for detailed values). MPS: mean patch size; MPE: mean patch edge; ED: edge density; LPI: largest patch index; SHDI: Shannon’s diversity index; SHEI: Shannon’s evenness index; MPAR: mean perimeter–area ratio; MSI: mean shape index; MFD: mean fractal dimension; NP: number of patches; MENND: mean Euclidean nearest neighbour distance; MPROX: mean proximity index.
Figure 5. Evolution of spatial metrics at the landscape level, for the period 1947–1979–2019 (see Table A1 for detailed values). MPS: mean patch size; MPE: mean patch edge; ED: edge density; LPI: largest patch index; SHDI: Shannon’s diversity index; SHEI: Shannon’s evenness index; MPAR: mean perimeter–area ratio; MSI: mean shape index; MFD: mean fractal dimension; NP: number of patches; MENND: mean Euclidean nearest neighbour distance; MPROX: mean proximity index.
Land 11 01828 g005
Land 11 01828 g006 550
Figure 6. Evolution of spatial metrics at the class level, for the period 1947–1979–2019 (see Table A2, Table A3 and Table A4 for detailed values). CA: total class area; CP: class proportion; LPI: largest patch index; MPS: mean patch size; MPE: mean patch edge; ED: edge density; MPAR: mean perimeter–area ratio; MSI: mean shape index; MFD: mean fractal dimension; MENND: mean Euclidean nearest neighbour distance; MPROX: mean proximity index; DIV: division index.
Figure 6. Evolution of spatial metrics at the class level, for the period 1947–1979–2019 (see Table A2, Table A3 and Table A4 for detailed values). CA: total class area; CP: class proportion; LPI: largest patch index; MPS: mean patch size; MPE: mean patch edge; ED: edge density; MPAR: mean perimeter–area ratio; MSI: mean shape index; MFD: mean fractal dimension; MENND: mean Euclidean nearest neighbour distance; MPROX: mean proximity index; DIV: division index.
Land 11 01828 g006
Land 11 01828 g007 550
Figure 7. Different sections of the city, showing the main tendencies of land cover evolution: (a) section of the city near the centre, where ABE cover has been dominant since at least 1947; (b) section of the city near the northern border, that faced a significant increase in ABE cover and decrease in HER cover; (c) section of the city in the western extremity, where the increase in ABE cover resulted in the reduction and fragmentation of HER and TRS cover; (d) section of the city near the eastern limit, where HER cover was replaced by TRS and ABE cover.
Figure 7. Different sections of the city, showing the main tendencies of land cover evolution: (a) section of the city near the centre, where ABE cover has been dominant since at least 1947; (b) section of the city near the northern border, that faced a significant increase in ABE cover and decrease in HER cover; (c) section of the city in the western extremity, where the increase in ABE cover resulted in the reduction and fragmentation of HER and TRS cover; (d) section of the city near the eastern limit, where HER cover was replaced by TRS and ABE cover.
Land 11 01828 g007
Table
Table 1. Land cover categories (adapted from [39]).
Table 1. Land cover categories (adapted from [39]).
Land Cover CategoriesAbbreviationDescription
Artificial Built ElementsABEImpervious surfaces, buildings, and other constructed elements.
Trees and ShrubsTRSWoody vegetation, including all phanerophytes and shrubby chamaephytes.
HerbaceousHERHerbaceous vegetation, including hemicryptophytes, therophytes and geophytes.
Sparsely Vegetated—TerrestrialSPVEvery type of non-vegetated soil, such as bare soil, sand, and rock.
Sparsely Vegetated—AquaticAQUEvery type of water surface, such as the sea, rivers, lakes, and ponds (including artificial elements).
Table
Table 2. Landscape metrics selected for this study (adapted from [52,53]).
Table 2. Landscape metrics selected for this study (adapted from [52,53]).
LevelTypeMetric
LandscapeArea and edge metricsTA: Total area (ha)
LPI: Largest patch index (%)
MPS: Mean patch size (ha)
MPE: Mean patch edge (m)
ED: Edge density (m/ha)
Shape metricsMPAR: Mean perimeter–area ratio
MSI: Mean shape index
MFD: Mean fractal dimension
Aggregation metricsNP: Number of patches
MENND: Mean Euclidean nearest neighbour distance (m)
MPROX: Mean proximity index
Diversity metricsCR: Class richness
SHDI: Shannon’s diversity index
SHEI: Shannon’s evenness index
ClassArea and edge metricsCA: Total class area (ha)
CP: Class proportion (%)
LPI: Largest patch index (%)
MPS: Mean patch size (ha)
MPE: Mean patch edge (m)
ED: Edge density (m/ha)
Shape metricsMPAR: Mean perimeter–area ratio
MSI: Mean shape index
MFD: Mean fractal dimension
Aggregation metricsNP: Number of patches
MENND: Mean Euclidean nearest neighbour distance (m)
MPROX: Mean proximity index
DIV: Division index
PatchArea and edge metricsArea (ha)
Edge (m)
Shape metricsPerimeter–area ratio
Shape index
Fractal dimension
Aggregation metricsEuclidean nearest neighbour distance (m)
Proximity index
Table
Table 3. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis at the landscape level (N = 4411; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Table 3. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis at the landscape level (N = 4411; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Patch MetricsKruskal–Wallis TestDunn’s Test
Test Statisticd.f.Sig.Pairwise ComparisonsAdj. Sig.
Area53.0692<0.001 *1947–19790.052
1947–2019<0.001 *
1979–2019<0.001 *
Edge18.0472<0.001 *1947–19790.001 *
1947–2019<0.001 *
1979–20191.000
Perimeter–area ratio403.9102<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Shape index251.8292<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Fractal dimension467.6542<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Euclidean nearest neighbour distance0.80320.669(pairwise comparisons were not performed as the overall test does not show significant differences across samples)
Proximity index59.7162<0.001 *1947–19790.020
1947–2019<0.001 *
1979–2019<0.001 *
* Significant differences.
Table
Table 4. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis for the ABE class (N = 363; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Table 4. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis for the ABE class (N = 363; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Patch MetricsKruskal–Wallis TestDunn’s Test
Test Statisticd.f.Sig.Pairwise ComparisonsAdj. Sig.
Area1.01920.601(pairwise comparisons were not performed)
Edge3.73420.155(pairwise comparisons were not performed)
Perimeter–area ratio1.76320.414(pairwise comparisons were not performed)
Shape index1.93320.380(pairwise comparisons were not performed)
Fractal dimension2.42620.297(pairwise comparisons were not performed)
Euclidean nearest neighbour distance94.7032<0.001 *1947–19790.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Proximity index120.7502<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
* Significant differences.
Table
Table 5. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis for the HER class (N = 1564; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Table 5. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis for the HER class (N = 1564; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Patch MetricsKruskal–Wallis TestDunn’s Test
Test Statisticd.f.Sig.Pairwise ComparisonsAdj. Sig.
Area87.6082<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Edge6.62120.036 *1947–19791.000
1947–20190.044 *
1979–20190.212
Perimeter–area ratio236.7922<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Shape index63.7662<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Fractal dimension221.4902<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Euclidean nearest neighbour distance33.9852<0.001 *1947–19790.028 *
1947–2019<0.001 *
1979–20190.002 *
Proximity index59.7162<0.001 *1947–19790.020
1947–2019<0.001 *
1979–2019<0.001 *
* Significant differences.
Table
Table 6. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis for the TRS class (N = 2321; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Table 6. Outputs of the Kruskal–Wallis test (and post hoc pairwise comparisons) on patch metrics, considering the analysis for the TRS class (N = 2321; significance values for Dunn’s test were adjusted by the Bonferroni correction for multiple tests).
Patch MetricsKruskal–Wallis TestDunn’s Test
Test Statisticd.f.Sig.Pairwise ComparisonsAdj. Sig.
Area8.28020.016 *1947–19790.486 *
1947–20190.012 *
1979–20190.402 *
Edge56.5232<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–20190.017 *
Perimeter–area ratio319.7092<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Shape index302.8952<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Fractal dimension452.8962<0.001 *1947–1979<0.001 *
1947–2019<0.001 *
1979–2019<0.001 *
Euclidean nearest neighbour distance4.50820.105(pairwise comparisons were not performed)
Proximity index0.92420.630(pairwise comparisons were not performed)
* Significant differences.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Guilherme, F.; Moreno, E.G.; Gonçalves, J.A.; Carretero, M.A.; Farinha-Marques, P. Looking Closer at the Patterns of Land Cover in the City of Porto, Portugal, between 1947 and 2019—A Contribution for the Integration of Ecological Data in Spatial Planning. Land 2022, 11, 1828. https://doi.org/10.3390/land11101828

AMA Style

Guilherme F, Moreno EG, Gonçalves JA, Carretero MA, Farinha-Marques P. Looking Closer at the Patterns of Land Cover in the City of Porto, Portugal, between 1947 and 2019—A Contribution for the Integration of Ecological Data in Spatial Planning. Land. 2022; 11(10):1828. https://doi.org/10.3390/land11101828

Chicago/Turabian Style

Guilherme, Filipa, Eva García Moreno, José Alberto Gonçalves, Miguel A. Carretero, and Paulo Farinha-Marques. 2022. "Looking Closer at the Patterns of Land Cover in the City of Porto, Portugal, between 1947 and 2019—A Contribution for the Integration of Ecological Data in Spatial Planning" Land 11, no. 10: 1828. https://doi.org/10.3390/land11101828

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop