Two and a half centuries of land reclamation, intensification, and urbanization homogenized northern Belgium landscapes

Two and a half centuries of land reclamation, intensification, and urbanization homogenized northern Belgium landscapes | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Two and a half centuries of land reclamation, intensification, and urbanization homogenized northern Belgium landscapes Luc De Keersmaeker, Pieter Roggemans, Lien Poelmans, Frederik Priem, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5536645/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Jan, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract We quantified historical land-use with deep learning segmentation, applied to tiled historical maps, and identified 3 successive drivers of long-term (1774–2022) landscape transformation in northern Belgium (13,800 km 2 ). Between 1774 and 1873, land reclamation halved the area of natural and semi-natural land-use. Agricultural intensification was the main driver in the next time interval (1873–1969), as the area of grassland and orchard doubled at the expense of arable land. Urbanization marked the last time interval (1969–2022) and reduced agricultural land-use. The reclamation of fertile soils for agriculture and the shift of forests to sand soils previously covered by heathland first increased the association of land-use classes to soil groups. After 1873 this association progressively weakened by expansion of grasslands beyond valleys and polders and urbanization disregarding soils. A sharp rise of land-use interspersion indicated that landscape transformation culminated between 1873 and 1969 and resulted in the homogenization of previously distinct landscapes. Scientific community and society/Geography Earth and environmental sciences/Environmental social sciences Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Historical land-use and consecutive land-use changes through time, along with the spatial context in which these occurred, determine to a high degree the characteristics and functioning of present-day landscapes. For instance, the decline and fragmentation of previously extensive land-use patches explain biodiversity hosted by present-day landscapes 1 . By identifying and quantifying the legacy effects of past land-use, landscape managers can develop strategies for biodiversity conservation and ecosystem restoration 2 . Similarly, cultural landscapes with heritage values at risk of disappearance can be identified to set priorities for conservation 3 . Past land-use can also explain the quality and the evolution over time of the ecosystem services landscapes can provide 4 . Whereas knowledge of historical land-use and its changes is highly relevant for many research areas, the availability, quality, and characteristics of historical sources often restrict data collection and analyses 5 . Land-use changes over the past decades can be accurately quantified and evaluated using remote sensing data, i.e. aerial photographs and satellite images, even on a global scale 6 . For studies that cover a longer time span, e.g. centuries, the reconstructions of historical land-use relies on databases of historical documents (e.g. census data), historical maps or pictures, natural archives (e.g. pollen or charcoal), or multiple-source data that can feed reconstruction models 7 8 9 . Historical topographic maps represent invaluable archives offering detailed insights into landscape configurations and land-use dynamics of past centuries 10 11 . Until recently, manual digitizing to some extent was necessary to create area-wide maps of historical land-use. Studies that applied this labor-intensive methodology focused on study sites that covered relatively small areas 12 13 14 . Machine learning and deep learning image classification techniques developed for remote sensing, also named GeoAI, opened up new possibilities for the automated extraction of information enclosed by historical maps 15 . Recent studies demonstrated that GeoAI can generate maps of historical land-use by categorizing the legend of historical maps, to study long-term land-use changes (Table 1 ). Table 1 Comparison of recent land-use change studies that applied GeoAI (machine and deep learning) techniques to historical maps, for the covered area, the native scale of the oldest map, the studied time span and the extracted number of land-use (LU) classes. Country Area (km 2 ) Native Scale Time span (years) Time slices LU classes Reference Belgium 13,800 1:11,520 248 4 9 this study Denmark 595 1:20,000 138 2 10 64 Finland 900 1:20,000 57 4 5 65 France 100,000 1:86,400 238 2 4 66 Great Britain 209,331 1:63,360 75 2 5 67 Ireland 3,025 1:10,565 172 2 2 68 Sweden 170,763 1:100,000 100 2 1 69 For this work, it is a prerequisite to use high quality scans of historical maps that are accurately georeferenced. The assemblage of map tiles expands the spatial extent of the area to which GeoAI techniques can be applied (see, e.g. 16 ). But even then, the study of land-use changes based on historical maps faces tradeoffs determined by the map qualities, i.e. the native scale, the map extent, the geometric accuracy, or the quality of the drawing and the legend. GeoAI studies that cover an area of at least 10,000 km 2 in Britain, France and Sweden, included only 1 historical map with a native scale that is less than 1:50,000 and extracted no more than 5 land-use classes. By contrast, similar studies in Denmark, Finland and Ireland used maps with higher native scales of at least 1:20,000 and extracted up to 10 land-use classes, but covered a much smaller area (Table 1 ). We applied deep learning GeoAI to 3 tiled historical maps for area-wide semantic segmentation of land-use in northern Belgium, on the west side of the North European Plain (Fig. 1 a, 1 b). The oldest map series was already finished in 1774, before industrialization, at a detailed scale of 1:11,520. Approximately 30,000 control points were used to correct distortions and create a high-quality map tile 17 . Similar land-use segmentations were applied on tiled 19th and 20th century topographical maps that, opposed to the 18th century map, did not need distortion corrections as triangulation and leveling techniques were used at that time 18 . The high quality of the 3 tiled historical map series enabled us to discern 9 common land-use classes and study land-use change with a high spatial resolution, using a 10 x 10 m grid, and for a relatively large area of 13,800 km 2 (Table 1 ). Complemented with a present-day map this provided 4 time slices that span 248 years, which is a solid basis for a comprehensive insight into land-use change since the late 18th century, in relation to a soil zonation that is characteristic of the North European Plain (Fig. 1 c, 1 d). Results Land-use change between 1774 and 2022 The land-use on the 1774, 1873, and 1969 maps was detected with overall accuracies of 94%, 91% and 93%, respectively. The thematic validation of the 2022 land-use map using a field survey indicated an overall accuracy of 86%. High values for both the user’s and producer’s accuracies of land-use classes with high area proportions (Supplementary Tables S1-4), attribute to the high overall accuracies. Whereas user’s accuracies of the segmented historical maps are consistently high, producer’s accuracies are relatively low for orchard (1774, 1873), water (1774), built-up & garden (1873), the intertidal zone (1969) and marshland (1969) (Supplementary Tables S1-3). The areas of these land-use classes are thus underestimated by the maps, but corrected by the area estimates (Table 2 ). Table 2 Area proportion estimates (%) with 95% confidence intervals (CI) of 9 land-use classes at 4 time slices in northern Belgium. The 1774, 1873, and 1969 area proportion estimates and CI validate the GeoAI segmentation of historical maps (Supplementary Tables S1, S2, S3, respectively). The area estimates and CI of the 2022 land-use map are validated with the LUCAS field survey data for 2021 61 . The area of heathland & dune in 2022 is not validated (NV), as the LUCAS typology does not discern this land-use class (Supplementary Table S4). Land-use class 1774 1873 1969 2022 Area (%) CI (%) Area (%) CI (%) Area (%) CI (%) Area (%) CI (%) Arable land 51.3 1.9 55.2 3.2 36.4 1.4 27.3 0.8 Built-up & garden 6.4 1.2 11.5 2.8 16.6 1.2 29.3 0.8 Forest 10.2 0.5 9.3 0.4 9.5 0.2 12.7 0.7 Freshwater marsh 0.8 0.1 0.3 0.1 0.2 0.1 0.3 0.2 Grassland 11.5 0.9 12.4 0.7 26.9 1.0 21.8 0.9 Heathland & dune 11.8 0.3 6.4 1.2 3.1 0.3 1.4 NV Intertidal zone 0.5 0.1 0.3 0.0 0.4 0.3 0.1 0.1 Orchard 3.6 1.6 2.1 1.2 4.7 0.9 1.8 0.3 Water 2.8 0.7 2.4 0.4 2.2 0.6 2.4 0.4 Mapped area 99.0 99.9 100.0 97.1 Not specified 1.0 0.1 0.0 2.9 Whereas the area of surface water was always between 2% and 3% of northern Belgium, area proportions of all other land-use classes changed between 1774 and 2022 (Table 2 ). The area of heathland & dune occupied an estimated 11.8% of the study area in 1774, mostly in the Northeast (Fig. 2 ), but was reduced by half in 1873 (Table 2 ). This land-use continued to decline after 1873, to 1.4% of the total area in 2022. Freshwater marsh (0.8%) and the intertidal zone (0.5%) covered much smaller areas in 1774, but both were also prone to a strong area decline between 1774 and 1873. The grassland area was stable (11.5–12.4%) in the first century and marked river valleys and polders on the 1774 and 1873 maps (Fig. 2 ). The grassland area more than doubled in the next time interval, up to 26.9% in 1969. This strong increase was followed by a decline to 21.8% in 2022 (Table 2 ). Arable land was the land-use class with the highest area proportion until 1969, but was surpassed by built-up area & garden in 2022. Whereas more than 50% of northern Belgium was used as arable land in the 18th and 19th century, this proportion declined to 36.4% in 1969 and 27.3% in 2022. The total forest area did not change much between 1774 and 1969 (9.3%-10.2%) and slightly increased to 12.7% in 2022. The area used for orchards amounted to 4.7% in 1969, but declined to 1.8% in 2022. Built-up area & garden continued to increase from 6.4% in 1774 to 29.3% in 2022, not only by expansion of historical cities and villages, but also by urban sprawl along connecting roads and dispersed small settlements (Fig. 2 ). Association of land-use with soil groups The V-score that represents the overall association of the nine land-use classes with seven soil groups, including not specified soils, first increased (1774–1873), but then declined after 1873 to values below that of 1774 (Fig. 3 ). This trend was mostly explained by the homogeneity, not so much by the completeness, of land-use classes in soil groups (Fig. 3 ). The increase of homogeneity indicates that soil groups contain a more equal distribution of land-use classes than before. The initial increase of the overall association is explained by the increase of arable land in the first time interval (1774–1873). This trend is seen not only on sandy loam, silt loam, and polder that are fertile, highly suitable soils for this land-use (Fig. 4a2, a3, a4), but also on less optimal sand and alluvial soil (Figs. 4a1, a6). By contrast, forest decreased on sandy loam, silt loam, and alluvial soil (Figs. 4c2, c3, c4). The initial increase of arable land in the polder area (Fig. 4a4) coincided with a decline of grassland (Fig. 4e4) and the intertidal zone (Fig. 4g4). The decline of forests on fertile sites (Figs. 4c2, c3, c4) and alluvial soil (Fig. 4c6) was compensated by an increase on sand soil (Fig. 4c1), where heathland & dune declined (Fig. 4f1). As a result the forest cover shifted from the South and West where silt loam and sandy loam soils dominate, to the sandy Northeast (Fig. 1 , 2 ). Between 1774 and 1873, heathland & dune also decreased on wet & organic soil (Fig. 4f5), where grassland further increased to approximately 70% (Fig. 4e5). The overall decrease of the association with soil groups in the next time interval (1873–1969) is explained by a shift of arable land (Fig. 4a1, a2, a3, a4, a6) to grassland (Fig. 4e1, e2, e3, e4, e6) at all soils except at wet & organic soil (Fig. 4a5, e5). Approximately two thirds of the grassland area was located on polder, wet & organic, and alluvial soil (Fig. 4e4, e5, e6) in 1774 and 1873, and approximately 25% on sand, sandy loam and silt loam soil (Fig. 4e1, e2, e3). In 1969 and 2022, this ratio was reversed. By contrast, the much smaller area of orchards contracted to silt loam soils after 1873 (Fig. 4h3). The strong increase of built-up area & garden and the decline of agricultural land, i.e. arable land, grassland, and orchard, accounted for the further weakening of the overall association between land-use and soil groups in the last time interval (1969–2022). Built-up area & garden increased at all soil groups, even on alluvial and wet & organic soils with a high risk of flooding (Fig. 4b1-7). The proportion of forested wet & organic soil approximately doubled after 1873 (Fig. 4c5), whereas the opposite trend was observed for grassland on this soil group (Fig. 4e5). Not specified soils by definition have become built-up through time, which opposes the trend of declining association between land-use classes and soil groups (Fig. 4b7). The proportion of not specified soils also increased in heathland & dune, as this land-use became scarce after 1873 and nowadays a considerable area is concentrated in military zones without soil classification. Changes of autocorrelation and interspersion Changes of spatial autocorrelation of land-use classes, expressed by global Moran’s I, mostly reflect the area changes (Fig. 5 ). The autocorrelation of arable land (Fig. 5 a), heathland & dune (Fig. 5 e), the intertidal zone (Fig. 5 f), and freshwater marshland (Fig. 5 g) decreased whereas the autocorrelation of built-up area & garden (Fig. 5 b) increased, in line with the area changes (Table 2 ). However, the area decline of heathland & dune, the intertidal zone, and freshwater marshland in the first time period (1774–1873) preceded the decline of autocorrelation that mainly occurred between 1873 and 1969 (compare Table 2 with Fig. 5 e, 5 f & 5 g). The area and autocorrelation of grassland followed opposite trends (compare Table 2 and Fig. 5 d). Although the grassland area doubled between 1774 and 2022, its dispersed increase on soil groups where it was previously scarce reduced spatial autocorrelation. The autocorrelation of orchards was higher in 2022 than before, in spite of the area decline in the last time interval (compare Table 2 and Fig. 5 h). The contraction of this land-use to silt loam soil is explanatory for this finding (Fig. 4 h). A complementary metric to evaluate the change of the landscape texture is the interspersion level, that refers to the spatial intermixing of different patch types, land-use classes in this case, without explicit reference to the dispersion level that we quantified by means of global Moran’s I. In 1774 and 1873, the interspersion level was less than 50% for approximately two thirds of northern Belgium (Fig. 6 ). River valleys, where several land-use classes converged, were accentuated by high interspersion levels, whereas vast areas of arable land, heathland & dune, and forest displayed the lowest interspersion levels (compare Fig. 2 and Fig. 7 ). Interspersion increased between 1774 and 1873 in the Northeast, as a consequence of the conversion of heathland & dune to forest and arable land (Fig. 7 ). Between 1873 and 1969 the interspersion of land-use classes increased sharply (Fig. 6 ) and masked the hydrological structure (Fig. 7 ), as a result of the dispersed increase of grassland and built-up area & garden (compare Fig. 2 and Fig. 7 ). Expanding cities, e.g. Brussels, Antwerp, Ghent and Bruges, have become islands of homogeneous land-use with little land-use interspersion in 1969 and 2022 (compare Fig. 2 and Fig. 7 ). Discussion Land reclamation The oldest map in our study, finished in 1774, depicts the Southern or Austrian Netherlands that belonged to the Habsburg Empire. The map is a snapshot of the landscape at the end of the Ancient Régime, before the socioeconomic changes inflicted by the French invasion in 1795 and industrialization transformed the landscape 17 . It is assumed that agricultural expansion was the primary response to population growth at that time causing the cropland area in most European countries to peak around 1900 19 . The land-use changes in northern Belgium in the first time interval (1774–1873) support this assumption. As a result heathland and dune vegetation, inland marshes, long-established forests, and the intertidal zone, i.e. natural and semi-natural land-use, declined in the same time interval. Heathland & dune vegetation and inland marshes, mostly located on sand soil, still covered more than 12% of northern Belgium in 1774. These were used as common land for pasturing, cutting of peat, digging of loam, and wood gathering. In Northwestern Europe the degradation of common land-use geared up in the last decades of the 18th century 20 . Within a century, the areas of heathland & dune, inland marshes, and the intertidal zone in northern Belgium were approximately reduced by half, and in the case of heathland & dune, the decline continued up to 2022. The shift of land-use on sand soil could indicate that forest plantation and reclamation to arable land respectively accounted for two thirds and one third of the heathland & dune area loss between 1774 and 1873. Our cartographic analysis also indicates that the area of inland marshes and the intertidal area was already reduced with 50% between 1774 and 1873, whereas the worldwide loss of wetlands is assumed to have largely taken place in the 20th century 21 . This finding confirms that the reclamation of marshes, i.e. wetlands with a permanent high water table, could have preceded the decline of other wetlands, e.g. wet grasslands 22 . Although the increasing availability of fossil pit coal fueled the early industrialization of Belgium from 1750 onward 23 , forests remained important for energy supply at least until the first decades of the 19th century 24 . However, in the aftermath of the French Revolution, forests of the clergy and aristocracy were confiscated, privatized, and converted to farmland 24 . Furthermore, a famine caused by the potato blight (1846–1850) exacerbated deforestation in Northwestern Belgium 25 . In many parts of Europe forest cover was at a minimum in the 19th century as a result of agricultural expansion 26 . The already low total forest cover in northern Belgium did not change much between 1774 and 1873, as the conversion of heathland and marshes to plantation forests with coniferous trees 24 compensated for the loss of 50% of the forest area on fertile soils. Low net forest area changes thus concealed the erosion of long-established ancient forests with a high naturalness 27 . Agricultural intensification European agriculture was challenged by globalization after 1870, in particular by the cheap import of overseas cereals 28 29 . At that time the import and industrial production of fertilizers also stimulated agricultural intensification 30 . Both facilitated a strong increase of livestock production at the end of the 19th century and even more in the 20th century 31 32 33 . In England 34 and the Netherlands 35 this caused a shift from cropland to pasture, and our study revealed that the grassland area doubled in northern Belgium between 1873 and 1969. By contrast, the increased focus on livestock production did not inflict such land-use change in Germany 36 , Spain 37 , Denmark 38 and Czech Republic 39 . As a result of intensification, Belgian agriculture not only shifted to livestock production, but also to horticulture and fruit production 29 . The area of orchards for fruit production increased between 1873 and 1969, and became concentrated on silt loam soils in the Southeast of our study area whereas dispersed farm-based orchards disappeared 14 . Urbanization The built-up & garden area increased from 6.4% in 1774 to 29.3% in 2022, which is more or less proportional to the population growth in that time interval, from between 1,500,000 and 2,000,000 inhabitants in 1774 to 8,000,000 in 2022. Population growth exceeded the area increase between 1774 and 1969, and can be explained by urban transition during industrialization in the 19th and the first half of the 20th centuries 40 . If we only consider the last time period (1969–2022), the area increase of built-up & garden (+ 76%) far exceeded population growth (+ 23%), which points to urban sprawl 41 . Our area-wide results confirm previous assessments of urban sprawl based on indirect population data at the level of municipalities 42 and case studies 43 . Urban sprawl in northern Belgium often manifests itself as ribbon development guided by historical patterns of roads, dikes and villages and is facilitated by a lack of a policy to prevent this self-reinforcing process 44 . As a result, this region is a hotspot of urbanization in recent decades 45 . The increase of infrastructure, buildings and gardens after 1969 was proportional on all soil groups, so polder, wet & organic, and alluvial soils were not avoided. This is a confirmation of case studies of river valleys 14 and the assessment that new settlements established in Belgium between 1985 and 2015 are evenly distributed over flooding risk categories 46 . Landscape transformation Landscape coherence, defined as the overall association between land-use and soil patterns 47 , first increased in northern Belgium between 1774 and 1873 as a result of land reclamation. In a similar way, by conversion of forest to arable land, the natural potential of fertile sites was further unleashed in 19th century Germany 48 and Poland 49 . By contrast, socioeconomic drivers explained land-use changes between 1860 and 1958 in a mountainous French Mediterranean landscape more than biophysical drivers did, but this was reversed afterwards 50 . As opposed to these areas in Germany, Poland, and France, that are still rural, the association between site conditions and land-use progressively weakened in northern Belgium after 1873. A similar trend occurred in the Netherlands between 1900 and 1990, also with an increase of grassland on a wide range of soil groups favoring this trend and a concentration of forest on coarse sand soil opposing it 35 . A study of regional differences at a single time slice in Estonia confirmed that land-use intensification increased the similarity between landscapes and thus the homogenization of the landscape structure 47 . The 3 drivers of land-use change, being land reclamation, agricultural intensification, and urbanization followed each other in time. The rise of the land-use interspersion level could indicate that landscape transformation culminated between 1873 and 1969. Up to 1873 land-use reflected the soil zonation and delineated distinct landscapes, e.g. grassland in valleys and polders, arable land on plateaus of sandy loam and silt loam, and also adjacent to villages on sand soils. These villages and surrounding arable lands on sand soil were separated by vast areas of heathland and marshes. From 1969 onwards, land-use is interspersed to a high degree, in particular by the grassland area that expanded beyond polders and valleys and by progressive urbanization that manifested itself as urban sprawl. As a result, northern Belgium transformed from a mostly rural region with distinct landscapes, to a homogenized peri-urban region. This trajectory could be illustrative for other peri-urban regions of Europe that show similar growth rates of the surface area (+ 78%) and the population (+ 33%) since the mid-1950’s 51 . Methods Study area and soil typology Northern Belgium, composed of the administrative Flemish and Brussels capital regions, is a flat or undulating region with an altitude below 290 m above the North Sea level (Fig. 1 a), at the southwest side of the North European Plain (Fig. 1 b). Topsoils of the North European Plain mainly consist of varying fractions of periglacial pleistocene aeolian sand and silt deposits 52 . As a result there is a north-south zonation of sand soil, sandy loam soil, and silt loam soil, also present in northern Belgium (Fig. 1 c). To determine the association between land-use and soils, the Belgian soil map was reclassified to 7 soil groups based on soil texture and drainage class that are robust soil properties determined by long-term pedogenic processes 53 . The Belgian soil map (1: 20,000) resulted from the National Soil Survey, carried out between 1947 and 1971. Sites classified as disturbed soil, built-up area, water, or not surveyed areas such as military zones, were grouped into a ‘not specified’ soil group. The climate of northern Belgium is temperate oceanic, as indicated by averages calculated for 1991–2020 54 . The annual temperature was 11.0°C, and the average daily maximum and minimum temperatures equaled 14.7°C and 7.3°C, respectively. On 189.8 days with precipitation per year, an average total precipitation of 837.1 mm was measured. From the North Sea shore in the West towards the eastern border, there is a slight increase in continentality. At the end of the 18th century the population of northern Belgium was estimated between 1,500,000 and 2,000,000 inhabitants 55 , increasing to 3,000,000 in 1873, and 6,500,000 in 1969. Presently there are 8,000,000 inhabitants 56 . Historical and present-day maps We used the maps of Ferraris that cover the Austrian Netherlands before the independence of Belgium, the first edition of topographical maps of Belgium by Dépôt de la Guerre and the topographical maps of Belgium by the Military Geographical Institute (Supplementary Table S5) for the segmentation of historical land-use. These 3 historical maps have a native scale of 1:11,520, 1:20,000 and 1:25,000, respectively, and are available as high-quality tiled maps at the Geopunt and Cartesius geoportals. The present-day land-use map of northern Belgium is composed of the map of land-use in 2022 of the region of Flanders 57 , complemented with a similar land-use map of the capital region of Brussels. The final year of the observations is used to refer to the maps, i.e. 1774, 1873, 1969, and 2022, respectively. The 1774 map does not completely cover the present-day territory of Belgium and not mapped communities along the present-day border (1.0% of the study area) are excluded from spatial analyses. Areas of the land-use 2022 map assigned to sports and recreation (2.9% of the study area) are also excluded, as they are not discerned on historical maps and presently consist of an unknown mixture of land-use, such as forest, grassland, water, buildings and infrastructure. Although the georeferencing of the tiled historical maps is of high quality 17 , positional errors of the order of 10s of meters are still likely. Positional errors are propagated in the overlay of time slices and influence landscapes indices calculated on it 58 . To avoid error propagation we did not analyze land-use conversions in a direct way, using an overlay of 4 time slices, but calculated proportions of land-use classes in soil groups and vice versa, to evaluate the shift of land-use through time in relation to soils. Land-use segmentation and validation We used OrthoSeg 59 , an open source GeoAI software package, for image segmentation applied to historical maps. The specific legend and drawing of each historical map series required a specific number of classes for segmentation, further called segmentation classes (Supplementary Table S5). The training was based on area-wide manual digitization of segmentation class polygons, in square boxes that measured 256 x 256 m for the 1774 and 1873 maps, and 128 x 128 m for the 1969 map. The prediction ran on downloaded 2,048 x 2,048 pixel map tiles, with a pixel resolution of 1 m for the 1774 and 1873 maps and 0.5 m for the 1969 map, and with an overlap of 128 pixels for the 1774 and 1873 maps, and 256 pixels for the 1969 map. A first run based on few and simple training data for the initially discerned segmentation classes was followed by a desktop evaluation of the outcome. Next additional training boxes were digitized to correct segmentation errors or remove gaps and thus improve the result of the following run. If necessary, additional segmentation classes were discerned and training data of the previous run were adjusted accordingly. This process was repeated until the desktop evaluation was favorable, and comprised 12, 13, and 6 iterations for the 1774, 1873, and 1969 maps, respectively (Supplementary Table S5). The output of the segmentation were polygon layers for each historical time slice, with 16 (1774), 19 (1873), and 29 (1969) segmentation classes (Supplementary Table S6). These segmentation classes were clustered to 9 common land-use classes and a class with not specified land-use (Supplementary Table S6), that also included areas of the 2022 map assigned to sports and recreation. The validation used area-based error matrices that take into account the selection bias introduced by the stratified sampling of the validation points, to calculate the estimated areas and confidence intervals (CI), and the producer’s, consumer’s and overall accuracies 60 (Supplementary Tables S1-4). For the validation of the segmented historical land-use maps we selected 50–100 points per segmentation class with randomly generated coordinates. The selected validation points were assigned to the discerned segmentation classes by visual interpretation of the historical map scans, in a QGIS environment. Next, the validation points were clustered to the 9 land-use classes in a similar way as the outcome of the segmentation (Supplementary Table S6). We used the 2021 European Land-Use/Land Cover Area Frame Statistical Survey (LUCAS) point data 61 to validate the thematic accuracy of the 2022 land-use map. The LUCAS validation points are assigned to harmonized land-use and land cover classes by means of a field survey, following a stratified random sampling on a 2 x 2 km grid covering Europe. The land-use class ‘heathland & dune’ in our study is assigned to shrubland in the LUCAS typology, whereas shrubland on historical maps was mostly assigned to forest based on the common green legend coloration. Therefore we did not include survey points of the LUCAS land cover categories D (Shrubland) and F (Bare soil and Lichens) (50 points) in the validation. Consequently, the area of land-use class heathland & dune was not validated or adjusted based on the area-based error matrix (Supplementary Table S4). The LUCAS category of ‘temporary grasslands’ (B55) included 180 validation points divided over our land-use classes ‘arable land’ (89 points) and ‘grassland’ (91 points). Both assessments were considered correct, as nowadays in our study area there is no clear distinction between temporary grassland, classified to LUCAS land cover category B (Cropland) and intensively used and fertilized permanent grassland, classified to LUCAS category E (Grassland). Spatial analyses We converted the polygon layers of historical land-use to grids with the same extent and 10 x 10 m cell size as the land-use 2022 map for spatial analyses on land-use classes: the association of land-use classes with soil groups, the spatial autocorrelation of land-use classes, and the interspersion of land-use classes. We analyzed the spatial association of 9 land-use classes with 7 soil groups at 4 time slices and used for this purpose the global V-measure, and its decomposition into the homogeneity and completeness measures 62 , of land-use classes in soil groups and vice versa. Calculations were performed in Python using the GDAL and Scikit-Learn packages. The four land-use grid layers were also used to evaluate the spatial autocorrelation of land-use classes, i.e., the degree of fragmentation or clustering of land-use. For this purpose we converted the grid layers to binary maps, for each land-use class at each time slice, and calculated the corresponding Moran’s I values 63 in Python using the GDAL, Numpy and SciPy packages. Incremental spatial lags, ranging between 10 m and 10,000 m, were used to assess spatial autocorrelation over various distances. The interspersion was calculated with the r.neighbors package of grass in QGIS on a focal area with 500 m radius. The interspersion of a cell is the percentage of cells containing values which differ from the values assigned to the center cell, in the circular neighborhood with a 500 m radius, plus 1. Declarations Data availability The tiled historical maps of 1774 can be viewed and accessed at the Geopunt regional geodata portal of Flanders: https://www.vlaanderen.be/datavindplaats/catalogus/ferraris-kaart-kabinetskaart-der-oostenrijkse-nederlanden-en-het-prinsbisdom-luik-1771-1778 Tiled historical topographical maps are online available at the federal Cartesius geodata portal: Topographical map of 1873: https://wmts.ngi.be/arcgis/rest/services/seamless_carto__default__3857__140/MapServer/tile/{z}/{y}/{x } Topographical map of 1969: https://wmts.ngi.be/arcgis/rest/services/seamless_carto__default__3857__1100/MapServer/tile/{z}/{y}/{x } The maps of segmented historical land-use can be viewed and accessed at: ttps:// www.vlaanderen.be/datavindplaats/catalogus/digitalisatie-historisch-landgebruik-en-landgebruiksveranderingen-in-vlaanderen-1778-2022 Code availability The applied OrthoSeg open source software package is developed for the segmentation of remote sensing images and maps shared by geoportals 59 and can be accessed at: https://github.com/orthoseg/orthoseg/wiki Competing interests The authors declare no competing interests. Author contributions Conceptualization: L.D.K., P.R., L.P. Methodology: P.R., L.D.K., L.P., F.P., S.T., T.P. Visualization: L.D.K., F.P. Funding acquisition: J.V.V. Writing: L.D.K., F.P., L.P., J.V.V., P.R. References Le Provost G et al (2020) Land-use history impacts functional diversity across multiple trophic groups. Proc. Natl. Acad. Sci. 117, 1573–1579 Foster D et al (2003) The Importance of Land-Use Legacies to Ecology and Conservation. Bioscience 53:77–88 Schulp CJE, Levers C, Kuemmerle T, Tieskens KF, Verburg PH (2019) Mapping and modelling past and future land use change in Europe’s cultural landscapes. Land Use Policy 80:332–344 Schirpke U et al (2023) Past and future impacts of land-use changes on ecosystem services in Austria. J Environ Manage 345:118728 Bürgi M, Östlund L, Mladenoff DJ (2017) Legacy Effects of Human Land Use: Ecosystems as Time-Lagged Systems. Ecosystems 20:94–103 Winkler K, Fuchs R, Rounsevell M, Herold M (2021) Global land use changes are four times greater than previously estimated. Nat Commun 12:2501 Cui Q et al (2014) Historical land-use and landscape change in southern Sweden and implications for present and future biodiversity. Ecol Evol 4:3555–3570 Earth Syst. Sci. Data 15, 1005–1035 (2023) Paprotny D, Mengel M (2023) Population, land use and economic exposure estimates for Europe at 100 m resolution from 1870 to 2020. Sci Data 10:372 Olah B (2009) Historical maps and their application in landscape ecological research. Ekologia 28:143–151 Loran C, Haegi S, Ginzler C (2018) Comparing historical and contemporary maps - a methodological framework for a cartographic map comparison applied to Swiss maps. Int J Geogr Inf Sci 32:2123–2139 Cousins SAO, Auffret AG, Lindgren J, Tränk L (2015) Regional-scale land-cover change during the 20th century and its consequences for biodiversity. Ambio 44:17–27 Picuno P, Cillis G, Statuto D (2019) Investigating the time evolution of a rural landscape: How historical maps may provide environmental information when processed using a GIS. Ecol Eng 139:105580 Lathouwers E, Segers Y, Verstraeten G (2023) Reconstructing valley landscapes. GIS-analyses of past land use changes in three Flemish river valleys since the late 18th century. Land Use Policy 135:106960 Kang Y, Gao S, Roth RE (2024) Artificial intelligence studies in cartography: a review and synthesis of methods, applications, and ethics. Cartogr Geogr Inf Sci 51:599–630 Ågren AM, Lin Y (2024) A fully automated model for land use classification from historical maps using machine learning. Remote Sens Appl Soc Environ 36:101349 Vervust S, Boùùaert MC, De Baets B, Van de Weghe N, De Maeyer P (2018) A Study of the Local Geometric Accuracy of Count de Ferraris’s Carte de cabinet (1770s) Using Differential Distortion Analysis. Cartogr J 55:16–35 Maeyer PD (2000) Base maps in Belgium. Belg Rev Belge Géographie 165–172. 10.4000/belgeo.13977 Ye Y et al (2023) Reconstruction of cropland change in European countries using integrated multisource data since AD 1800. Boreas 52:60–77 De Moor M, Shaw-Taylor L, Warde P (2002) Comparing the historical commons of north west Europe. An introduction. The management of common land in north west Europe, c. 1500–1850, vol 8. Brepols, pp 15–31 Fluet-Chouinard E et al (2023) Extensive global wetland loss over the past three centuries. Nature 614:281–286 Godet L, Thomas A (2013) Three centuries of land cover changes in the largest French Atlantic wetland provide new insights for wetland conservation. Appl Geogr 42:133–139 Ryckbosch W, Saelens W (2023) Fuelling the urban economy: A comparative study of energy in the Low Countries, 1600–1850. Econ Hist Rev 76:221–256 Tallier P-A, Verboven H, Vandekerkhove K, Baeté H, Verheyen K (2018) CHAPTER 3 State Forestry in Belgium since the End of the Eighteenth Century. In: Oosthoek KJ, Hölzl R (eds) Managing Northern Europe’s Forests: Histories from the Age of Improvement to the Age of Ecology. Berghahn Books, pp 92–129. doi: 10.1515/9781785336010-007 . Verheyen K, Bossuyt B, Hermy M, Tack G (1999) The land use history (1278–1990) of a mixed hardwood forest in western Belgium and its relationship with chemical soil characteristics. J Biogeogr 26:1115–1128 Kaplan JO, Krumhardt KM, Zimmermann N (2009) The prehistoric and preindustrial deforestation of Europe. Quat Sci Rev 28:3016–3034 Bergès L, Dupouey J-L (2021) Historical ecology and ancient forests: Progress, conservation issues and scientific prospects, with some examples from the French case. J Veg Sci 32:e12846 O’Rourke KH (1997) The European Grain Invasion, 1870–1913. J Econ Hist 57:775–801 Van Dijck M, Truyts T (2014) The Agricultural Invasion and the Political Economy of Agricultural Trade Policy in Belgium, 1875–1900. SSRN Scholarly Paper at https://doi.org/10.2139/ssrn.2378740 Strotmann C, Herment L, Page A (2021) Fertilisers in the Long 19th Century and Beyond: Usage, Commercialisation and Production (c 1800–1939). Jahrb. Für Wirtsch. Econ. Hist. Yearb. 62, 1–18 Federico G, THE GROWTH OF WORLD AGRICULTURAL, PRODUCTION (1938) 1800– in Research in Economic History vol. 22 125–181 (Emerald Group Publishing Limited, 2004) Swinnen JFM (2010) Agricultural Protection Growth in Europe, 1870–1969. In: Anderson K (ed) The Political Economy of Agricultural Price Distortions. Cambridge University Press, Cambridge, pp 141–161. doi: 10.1017/CBO9780511778964.008 . Ramankutty N et al (2018) Trends in Global Agricultural Land Use: Implications for Environmental Health and Food Security. Annu Rev Plant Biol 69:789–815 Howkins A (2021) Reshaping Rural England: A Social History 1850–1925. Routledge, London. 10.4324/9780203421185 Bakker MM, Sonneveld MPW, Brookhuis B, Kuhlman T (2013) Trends in soil–land-use relationships in the Netherlands between 1900 and 1990. Agric Ecosyst Environ 181:134–143 Niedertscheider M, Kuemmerle T, Müller D, Erb K-H (2014) Exploring the effects of drastic institutional and socio-economic changes on land system dynamics in Germany between 1883 and 2007. Glob Environ Change 28:98–108 de González M et al (2020) Agricultural Output: From Crop Specialization to Livestocking, 1900–2008. In: de González M et al (eds) The Social Metabolism of Spanish Agriculture, 1900–2008: The Mediterranean Way Towards Industrialization. Springer International Publishing, Cham, pp 29–68. doi: 10.1007/978-3-030-20900-1_2 . Lohrum N, Normand S, Dalgaard T, Graversgaard M (2024) Unveiling the frontiers: Historical expansion and modern implications of agricultural land use in Denmark. J Environ Manage 359:120934 Moravcova J, Moravcova V, Pavlicek T, Novakova N (2022) Land Use Has Changed through the Last 200 Years in Various Production Areas of South Bohemia. Land 11:1619 Bocquier P, Costa R (2015) Which transition comes first? Urban and demographic transitions in Belgium and Sweden. Demogr Res 33:1297–1332 Knaap G-J, Song Y, Nedovic-Budic Z (2007) Measuring Patterns of Urban Development: New Intelligence for the War on Sprawl. Local Environ 12:239–257 Mustafa A, Teller J (2020) Self-Reinforcing Processes Governing Urban Sprawl in Belgium: Evidence over Six Decades. Sustainability 12:4097 Poelmans L, Van Rompaey A (2009) Detecting and modelling spatial patterns of urban sprawl in highly fragmented areas: A case study in the Flanders–Brussels region. Landsc Urban Plan 93:10–19 Verbeek T, Boussauw K, Pisman A (2014) Presence and trends of linear sprawl: Explaining ribbon development in the north of Belgium. Landsc Urban Plan 128:48–59 Kuemmerle T et al (2016) Hotspots of land use change in Europe. Environ Res Lett 11:064020 Rapid Urban Growth in Flood Zones: Global Evidence since 1985 (2022) 10.1596/1813-9450-1014 Mander Ü, Uuemaa E, Roosaare J, Aunap R, Antrop M (2010) Coherence and fragmentation of landscape patterns as characterized by correlograms: A case study of Estonia. Landsc Urban Plan 94:31–37 Wulf M, Jahn U, Meier K (2016) Land cover composition determinants in the Uckermark (NE Germany) over a 220-year period. Reg Environ Change 16:1793–1805 Zgłobicki W, Gawrysiak L, Baran-Zgłobicka B, Telecka M (2016) Long-term forest cover changes, within an agricultural region, in relation to environmental variables, Lubelskie province, Eastern Poland. Environ Earth Sci 75:1373 Abadie J et al (2018) Forest recovery since 1860 in a Mediterranean region: drivers and implications for land use and land cover spatial distribution. Landsc Ecol 33:289–305 Urban sprawl in Europe - The ignored challenge. Eur Environ Agency https://www.eea.europa.eu/publications/eea_report_2006_10/eea_report_10_2006.pdf Bertran P et al (2021) Revised map of European aeolian deposits derived from soil texture data. Quat Sci Rev 266:107085 De Keersmaeker L et al (2013) Application of the Ancient Forest Concept to Potential Natural Vegetation Mapping in Flanders, A Strongly Altered Landscape in Northern Belgium. Folia Geobot 48:137–162 KMI - Jaar KMI https://www.meteo.be/nl/klimaat/klimaat-van-belgie/klimatologisch-overzicht/2023/jaar Klep PM (1991) Population Estimates of Belgium, by Province (1375–1831). Société Belge de Démographie, Louvain-la-Neuve Key Figs (2023) | Statbel. https://statbel.fgov.be/en/news/key-figures-2023 Poelmans L, Janssen L, Hambsch L (2023) Landgebruik en ruimtebeslag in Vlaanderen, toestand 2022 - Onderzoeksportaal. https://www.friscris.be/nl/publications/landgebruik-en-ruimtebeslag-in-vlaanderen-toestand-2022(ab4c67af-a12e-4acf-9346-39df99589565).html Burnicki AC (2012) Impact of error on landscape pattern analyses performed on land-cover change maps. Landsc Ecol 27:713–729 Roggemans P (2024) orthoseg. Zenodo https://doi.org/10.5281/ZENODO.10340584 Olofsson P, Foody GM, Stehman SV, Woodcock CE (2013) Making better use of accuracy data in land change studies: Estimating accuracy and area and quantifying uncertainty using stratified estimation. Remote Sens Environ 129:122–131 The thematic accuracy of Corine land cover 2000 - Assessment using LUCAS. Eur Environ Agency https://www.eea.europa.eu/publications/technical_report_2006_7 Nowosad J, Stepinski TF (2018) Spatial association between regionalizations using the information-theoretical V -measure Moran PAP (1950) Notes on Continuous Stochastic Phenomena. Biometrika 37:17–23 Levin G, Groom G, Svenningsen SR (2024) Assessing long-term landscape dynamics based on automated production of land category layers from Danish late 19th century topographic maps. Preprint at. https://doi.org/10.21203/rs.3.rs-4021413/v1 Mäyrä J, Kivinen S, Keski-Saari S, Poikolainen L, Kumpula T (2023) Utilizing historical maps in identification of long-term land use and land cover changes. Ambio 52:1777–1792 Martinez T et al (2023) Deep learning ancient map segmentation to assess historical landscape changes. J Maps 19:2225071 Suggitt AJ et al (2023) Linking climate warming and land conversion to species’ range changes across Great Britain. Nat Commun 14:6759 O’Hara R, Marwaha R, Zimmermann J, Saunders M, Green S (2024) Unleashing the power of old maps: Extracting symbology from nineteenth century maps using convolutional neural networks to quantify modern land use on historic wetlands. Ecol Indic 158:111363 Ståhl N, Weimann L (2022) Identifying wetland areas in historical maps using deep convolutional neural networks. Ecol Inf 68:101557 Euro Dem | Eurogeographics https://www.mapsforeurope.org/datasets/euro-dem Additional Declarations There is NO Competing Interest. Supplementary Files LandscapetransformationNCsupplementary.docx Dataset 1 Cite Share Download PDF Status: Published Journal Publication published 21 Jan, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5536645","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":400155756,"identity":"dd15b681-1fc9-4a49-8035-eb7f19f3298a","order_by":0,"name":"Luc De Keersmaeker","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8UlEQVRIiWNgGAWjYBACPhDBw5AA4wMZ7I0NzBAWdsCGqYXnIMlaJBIY8GuRSH7A8KYiTY5f+vjTDT+ADPOZj9seF1Qw5PHj1JJmwDjnTI6xZF+O2c0eIEPmdmK78YwzDMWSDbi05DAw87ZVJG44w8N2A8SYIZ3YJs3bxpC44QA+Lf8q6vefYX928y+QMUPyIETLfrxaGnISDHgYzG6DGBISjFBbcPmF55nBwTnH0gxnnOExuy0DYvAAHTbjjESxBA5b+NmTHz54U5Msz98DdBiIIcF+/Jl0QYVNHj8O74MAVtMkcKsfBaNgFIyCUUAQAAD6T1XmEYI54AAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-9709-512X","institution":"Research Institute for Nature and Forest","correspondingAuthor":true,"prefix":"","firstName":"Luc","middleName":"","lastName":"De Keersmaeker","suffix":""},{"id":400155757,"identity":"947cf4ec-c678-4e21-902c-8c897b96180c","order_by":1,"name":"Pieter Roggemans","email":"","orcid":"https://orcid.org/0009-0009-2046-3284","institution":"Agency for Agriculture and Fisheries","correspondingAuthor":false,"prefix":"","firstName":"Pieter","middleName":"","lastName":"Roggemans","suffix":""},{"id":400155758,"identity":"8aec8933-4612-41cf-a3da-e639a59502f3","order_by":2,"name":"Lien Poelmans","email":"","orcid":"https://orcid.org/0000-0002-0716-6594","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Lien","middleName":"","lastName":"Poelmans","suffix":""},{"id":400155759,"identity":"6238ef4b-c7c5-4b17-8a69-2221b783b852","order_by":3,"name":"Frederik Priem","email":"","orcid":"","institution":"Flemish Institute for Technological Research (VITO NV)","correspondingAuthor":false,"prefix":"","firstName":"Frederik","middleName":"","lastName":"Priem","suffix":""},{"id":400155760,"identity":"129a5e40-d7ba-4439-a098-fc81f464eff6","order_by":4,"name":"Stijn Taillir","email":"","orcid":"","institution":"Digital Flanders Agency","correspondingAuthor":false,"prefix":"","firstName":"Stijn","middleName":"","lastName":"Taillir","suffix":""},{"id":400155761,"identity":"5a34581f-9040-42a2-aa6f-5a4bb09bcd81","order_by":5,"name":"Toon Petermans","email":"","orcid":"","institution":"Digital Flanders Agency","correspondingAuthor":false,"prefix":"","firstName":"Toon","middleName":"","lastName":"Petermans","suffix":""},{"id":400155762,"identity":"4ffa4dac-ca5a-41e7-912f-1812250a3968","order_by":6,"name":"Jo Van Valckenborgh","email":"","orcid":"","institution":"Digital Flanders Agency","correspondingAuthor":false,"prefix":"","firstName":"Jo","middleName":"Van","lastName":"Valckenborgh","suffix":""}],"badges":[],"createdAt":"2024-11-27 15:25:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5536645/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5536645/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-026-68594-y","type":"published","date":"2026-01-21T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":73469192,"identity":"faa87519-f96c-4ec1-b416-c35d37a28e52","added_by":"auto","created_at":"2025-01-10 09:22:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1958077,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNorthern Belgium is located at the west side of the North European Plain with a characteristic silt loam to sand soil zonation.\u003c/strong\u003e Northern Belgium is a flat or undulating region with an altitude below 300 m above North sea level (\u003cstrong\u003ea\u003c/strong\u003e), at the west side of the North European Plain (\u003cstrong\u003eb\u003c/strong\u003e), as shown by the digital elevation model \u003csup\u003e70\u003c/sup\u003e. The soil map of northern Belgium (\u003cstrong\u003ec\u003c/strong\u003e)\u0026nbsp; displays a north-south zonation from sand to silt loam soils, intersected by river valleys marked by alluvial, wet and organic soils. The area of the white box (\u003cstrong\u003ec\u003c/strong\u003e) is displayed enlarged (\u003cstrong\u003ed\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/8d95957a90cb66266c64e89f.png"},{"id":73469191,"identity":"02e7f06b-800f-4d13-8b34-59eec265d0e8","added_by":"auto","created_at":"2025-01-10 09:22:19","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":6519157,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLand-use change in northern Belgium between 1774 and 2022.\u003c/strong\u003e Land-use was mapped in 9 classes using a grid with 10 x 10 m resolution for spatial analyses. The area of the white box at the transition of silt loam to sand soils (see also Fig. 1d) is enlarged at the right side.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/32fd9652b0d7ba5abefc2255.png"},{"id":73469190,"identity":"a007589e-0bd8-4308-ad63-870dd3880cb0","added_by":"auto","created_at":"2025-01-10 09:22:19","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":260684,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe change of the association between 9 land-use (LU) classes and 7 soil groups in northern Belgium, from 1774 to 2022.\u003c/strong\u003e The homogeneity of land-use in soils measures how much of a soil group is composed of the same land-use class. The completeness measures how much of a land-use class is located within a soil group. The V-score or global association simultaneously assesses homogeneity and completeness. The measures are calculated on 10 x 10 m grids of the soil map (Fig 1c) and land-use maps at 4 time slices (Fig 2).\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/fdd246a7bc8f2a3871610c65.png"},{"id":73469203,"identity":"a66d94ee-587b-447a-b37b-410f91ec5b9d","added_by":"auto","created_at":"2025-01-10 09:22:20","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":976399,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe change of proportions (%) of land-use (LU) in soil groups and vice versa in northern Belgium, between 1774 and 2022.\u003c/strong\u003e The sum of 9 land-use classes ordered in rows in a soil group (blue bars), as well as the sum of 7 soil groups ordered in columns in a land-use class (orange bars), both equal 100%.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/e0b0639b624502570da3566b.png"},{"id":73470679,"identity":"7cfb1a51-2f99-4818-9fe2-f5f5a3232273","added_by":"auto","created_at":"2025-01-10 09:30:20","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":456647,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChange of spatial autocorrelation of land-use classes in northern Belgium, between 1774 and 2022. \u003c/strong\u003eGlobal Moran’s I value is calculated as a measure of spatial autocorrelation on 10 x 10 m binary grids of land-use, using an annular kernel with a variable radius or spatial lag (x-axis). A value of 1 (y-axis) indicates perfect clustering of a land-use class over the evaluated distance, and declining values indicate decreasing consistency. Negative Moran’s I values down to -1, that would indicate self-repulsion, were not observed.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/6c2c2499be9d0c501b6a936e.png"},{"id":73469208,"identity":"8a9a9ce0-3e4d-4265-a6f6-34ced036deb0","added_by":"auto","created_at":"2025-01-10 09:22:20","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":586251,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChange of the cumulative distribution of the interspersion level in northern Belgium, between 1774 and 2022.\u003c/strong\u003e The interspersion level is the percentage plus one of cells with land-use different from the focal raster cell with 10 x 10 m dimensions, within a circular kernel with a radius of 500 m.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/f629d5cf6b155871aeee5dae.png"},{"id":73469213,"identity":"a3a206f2-5421-4376-b769-8310b8a03201","added_by":"auto","created_at":"2025-01-10 09:22:20","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":6581531,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChange of the interspersion level in northern Belgium between 1774 and 2022.\u003c/strong\u003eThe interspersion level is the percentage plus one of cells with land-use different from the focal raster cell with 10 x 10 m dimensions, within a circular kernel with a radius of 500 m. The area of the white box at the transition of silt loam to sand soils (see also Fig. 1d, 2) is enlarged at the right side.\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/60f6e98f213fa54626a2a232.png"},{"id":103119543,"identity":"c6f0121a-c21e-4b91-abf9-5a8daac8c7c2","added_by":"auto","created_at":"2026-02-21 08:11:25","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":21427875,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/a80758ec-26e3-4ed3-9dec-1104e3e3c05d.pdf"},{"id":73470672,"identity":"e1499e0e-6cda-418c-88cd-a37e19984cba","added_by":"auto","created_at":"2025-01-10 09:30:19","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2115955,"visible":true,"origin":"","legend":"Dataset 1","description":"","filename":"LandscapetransformationNCsupplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-5536645/v1/d9e5c8469ba66c1ebdc1df64.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Two and a half centuries of land reclamation, intensification, and urbanization homogenized northern Belgium landscapes","fulltext":[{"header":"Introduction","content":"\u003cp\u003eHistorical land-use and consecutive land-use changes through time, along with the spatial context in which these occurred, determine to a high degree the characteristics and functioning of present-day landscapes. For instance, the decline and fragmentation of previously extensive land-use patches explain biodiversity hosted by present-day landscapes \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. By identifying and quantifying the legacy effects of past land-use, landscape managers can develop strategies for biodiversity conservation and ecosystem restoration \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Similarly, cultural landscapes with heritage values at risk of disappearance can be identified to set priorities for conservation \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Past land-use can also explain the quality and the evolution over time of the ecosystem services landscapes can provide \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWhereas knowledge of historical land-use and its changes is highly relevant for many research areas, the availability, quality, and characteristics of historical sources often restrict data collection and analyses \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Land-use changes over the past decades can be accurately quantified and evaluated using remote sensing data, i.e. aerial photographs and satellite images, even on a global scale \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. For studies that cover a longer time span, e.g. centuries, the reconstructions of historical land-use relies on databases of historical documents (e.g. census data), historical maps or pictures, natural archives (e.g. pollen or charcoal), or multiple-source data that can feed reconstruction models \u003csup\u003e7 8 9\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHistorical topographic maps represent invaluable archives offering detailed insights into landscape configurations and land-use dynamics of past centuries \u003csup\u003e10 11\u003c/sup\u003e. Until recently, manual digitizing to some extent was necessary to create area-wide maps of historical land-use. Studies that applied this labor-intensive methodology focused on study sites that covered relatively small areas \u003csup\u003e12 13 14\u003c/sup\u003e. Machine learning and deep learning image classification techniques developed for remote sensing, also named GeoAI, opened up new possibilities for the automated extraction of information enclosed by historical maps \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Recent studies demonstrated that GeoAI can generate maps of historical land-use by categorizing the legend of historical maps, to study long-term land-use changes (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparison of recent land-use change studies that applied GeoAI (machine and deep learning) techniques to historical maps, for the covered area, the native scale of the oldest map, the studied time span and the extracted number of land-use (LU) classes.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCountry\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea\u003c/p\u003e \u003cp\u003e(km\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNative\u003c/p\u003e \u003cp\u003eScale\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTime span\u003c/p\u003e \u003cp\u003e(years)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTime\u003c/p\u003e \u003cp\u003eslices\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eLU classes\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eReference\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBelgium\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13,800\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1:11,520\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e248\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003ethis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDenmark\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e595\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1:20,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e138\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFinland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e900\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1:20,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFrance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e100,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1:86,400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e238\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGreat Britain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e209,331\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1:63,360\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIreland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3,025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1:10,565\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e172\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSweden\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e170,763\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1:100,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eFor this work, it is a prerequisite to use high quality scans of historical maps that are accurately georeferenced. The assemblage of map tiles expands the spatial extent of the area to which GeoAI techniques can be applied (see, e.g. \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e). But even then, the study of land-use changes based on historical maps faces tradeoffs determined by the map qualities, i.e. the native scale, the map extent, the geometric accuracy, or the quality of the drawing and the legend. GeoAI studies that cover an area of at least 10,000 km\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e in Britain, France and Sweden, included only 1 historical map with a native scale that is less than 1:50,000 and extracted no more than 5 land-use classes. By contrast, similar studies in Denmark, Finland and Ireland used maps with higher native scales of at least 1:20,000 and extracted up to 10 land-use classes, but covered a much smaller area (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe applied deep learning GeoAI to 3 tiled historical maps for area-wide semantic segmentation of land-use in northern Belgium, on the west side of the North European Plain (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). The oldest map series was already finished in 1774, before industrialization, at a detailed scale of 1:11,520. Approximately 30,000 control points were used to correct distortions and create a high-quality map tile \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Similar land-use segmentations were applied on tiled 19th and 20th century topographical maps that, opposed to the 18th century map, did not need distortion corrections as triangulation and leveling techniques were used at that time \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. The high quality of the 3 tiled historical map series enabled us to discern 9 common land-use classes and study land-use change with a high spatial resolution, using a 10 x 10 m grid, and for a relatively large area of 13,800 km\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Complemented with a present-day map this provided 4 time slices that span 248 years, which is a solid basis for a comprehensive insight into land-use change since the late 18th century, in relation to a soil zonation that is characteristic of the North European Plain (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eLand-use change between 1774 and 2022\u003c/h2\u003e \u003cp\u003eThe land-use on the 1774, 1873, and 1969 maps was detected with overall accuracies of 94%, 91% and 93%, respectively. The thematic validation of the 2022 land-use map using a field survey indicated an overall accuracy of 86%. High values for both the user\u0026rsquo;s and producer\u0026rsquo;s accuracies of land-use classes with high area proportions (Supplementary Tables S1-4), attribute to the high overall accuracies. Whereas user\u0026rsquo;s accuracies of the segmented historical maps are consistently high, producer\u0026rsquo;s accuracies are relatively low for orchard (1774, 1873), water (1774), built-up \u0026amp; garden (1873), the intertidal zone (1969) and marshland (1969) (Supplementary Tables S1-3). The areas of these land-use classes are thus underestimated by the maps, but corrected by the area estimates (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eArea proportion estimates (%) with 95% confidence intervals (CI) of 9 land-use classes at 4 time slices in northern Belgium. The 1774, 1873, and 1969 area proportion estimates and CI validate the GeoAI segmentation of historical maps (Supplementary Tables S1, S2, S3, respectively). The area estimates and CI of the 2022 land-use map are validated with the LUCAS field survey data for 2021 \u003csup\u003e61\u003c/sup\u003e. The area of heathland \u0026amp; dune in 2022 is not validated (NV), as the LUCAS typology does not discern this land-use class (Supplementary Table S4).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLand-use class\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e1774\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003e1873\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e1969\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e \u003cp\u003e2022\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCI\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eArea (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCI\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eArea (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eCI\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eArea (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eCI\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArable land\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e55.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e36.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e27.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBuilt-up \u0026amp; garden\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e16.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e29.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e9.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e12.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFreshwater marsh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrassland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e12.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e26.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e21.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeathland \u0026amp; dune\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e3.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eNV\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntertidal zone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOrchard\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e4.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMapped area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e99.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e99.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e100.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e97.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot specified\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eWhereas the area of surface water was always between 2% and 3% of northern Belgium, area proportions of all other land-use classes changed between 1774 and 2022 (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The area of heathland \u0026amp; dune occupied an estimated 11.8% of the study area in 1774, mostly in the Northeast (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), but was reduced by half in 1873 (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). This land-use continued to decline after 1873, to 1.4% of the total area in 2022. Freshwater marsh (0.8%) and the intertidal zone (0.5%) covered much smaller areas in 1774, but both were also prone to a strong area decline between 1774 and 1873. The grassland area was stable (11.5\u0026ndash;12.4%) in the first century and marked river valleys and polders on the 1774 and 1873 maps (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The grassland area more than doubled in the next time interval, up to 26.9% in 1969. This strong increase was followed by a decline to 21.8% in 2022 (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Arable land was the land-use class with the highest area proportion until 1969, but was surpassed by built-up area \u0026amp; garden in 2022. Whereas more than 50% of northern Belgium was used as arable land in the 18th and 19th century, this proportion declined to 36.4% in 1969 and 27.3% in 2022. The total forest area did not change much between 1774 and 1969 (9.3%-10.2%) and slightly increased to 12.7% in 2022. The area used for orchards amounted to 4.7% in 1969, but declined to 1.8% in 2022. Built-up area \u0026amp; garden continued to increase from 6.4% in 1774 to 29.3% in 2022, not only by expansion of historical cities and villages, but also by urban sprawl along connecting roads and dispersed small settlements (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAssociation of land-use with soil groups\u003c/h3\u003e\n\u003cp\u003eThe V-score that represents the overall association of the nine land-use classes with seven soil groups, including not specified soils, first increased (1774\u0026ndash;1873), but then declined after 1873 to values below that of 1774 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). This trend was mostly explained by the homogeneity, not so much by the completeness, of land-use classes in soil groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The increase of homogeneity indicates that soil groups contain a more equal distribution of land-use classes than before.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe initial increase of the overall association is explained by the increase of arable land in the first time interval (1774\u0026ndash;1873). This trend is seen not only on sandy loam, silt loam, and polder that are fertile, highly suitable soils for this land-use (Fig.\u0026nbsp;4a2, a3, a4), but also on less optimal sand and alluvial soil (Figs.\u0026nbsp;4a1, a6). By contrast, forest decreased on sandy loam, silt loam, and alluvial soil (Figs.\u0026nbsp;4c2, c3, c4). The initial increase of arable land in the polder area (Fig.\u0026nbsp;4a4) coincided with a decline of grassland (Fig.\u0026nbsp;4e4) and the intertidal zone (Fig.\u0026nbsp;4g4). The decline of forests on fertile sites (Figs.\u0026nbsp;4c2, c3, c4) and alluvial soil (Fig.\u0026nbsp;4c6) was compensated by an increase on sand soil (Fig.\u0026nbsp;4c1), where heathland \u0026amp; dune declined (Fig.\u0026nbsp;4f1). As a result the forest cover shifted from the South and West where silt loam and sandy loam soils dominate, to the sandy Northeast (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Between 1774 and 1873, heathland \u0026amp; dune also decreased on wet \u0026amp; organic soil (Fig.\u0026nbsp;4f5), where grassland further increased to approximately 70% (Fig.\u0026nbsp;4e5).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe overall decrease of the association with soil groups in the next time interval (1873\u0026ndash;1969) is explained by a shift of arable land (Fig.\u0026nbsp;4a1, a2, a3, a4, a6) to grassland (Fig.\u0026nbsp;4e1, e2, e3, e4, e6) at all soils except at wet \u0026amp; organic soil (Fig.\u0026nbsp;4a5, e5). Approximately two thirds of the grassland area was located on polder, wet \u0026amp; organic, and alluvial soil (Fig.\u0026nbsp;4e4, e5, e6) in 1774 and 1873, and approximately 25% on sand, sandy loam and silt loam soil (Fig.\u0026nbsp;4e1, e2, e3). In 1969 and 2022, this ratio was reversed. By contrast, the much smaller area of orchards contracted to silt loam soils after 1873 (Fig.\u0026nbsp;4h3).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe strong increase of built-up area \u0026amp; garden and the decline of agricultural land, i.e. arable land, grassland, and orchard, accounted for the further weakening of the overall association between land-use and soil groups in the last time interval (1969\u0026ndash;2022). Built-up area \u0026amp; garden increased at all soil groups, even on alluvial and wet \u0026amp; organic soils with a high risk of flooding (Fig.\u0026nbsp;4b1-7). The proportion of forested wet \u0026amp; organic soil approximately doubled after 1873 (Fig.\u0026nbsp;4c5), whereas the opposite trend was observed for grassland on this soil group (Fig.\u0026nbsp;4e5).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNot specified soils by definition have become built-up through time, which opposes the trend of declining association between land-use classes and soil groups (Fig.\u0026nbsp;4b7). The proportion of not specified soils also increased in heathland \u0026amp; dune, as this land-use became scarce after 1873 and nowadays a considerable area is concentrated in military zones without soil classification.\u003c/p\u003e\n\u003ch3\u003eChanges of autocorrelation and interspersion\u003c/h3\u003e\n\u003cp\u003eChanges of spatial autocorrelation of land-use classes, expressed by global Moran\u0026rsquo;s I, mostly reflect the area changes (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The autocorrelation of arable land (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ea), heathland \u0026amp; dune (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ee), the intertidal zone (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ef), and freshwater marshland (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eg) decreased whereas the autocorrelation of built-up area \u0026amp; garden (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eb) increased, in line with the area changes (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). However, the area decline of heathland \u0026amp; dune, the intertidal zone, and freshwater marshland in the first time period (1774\u0026ndash;1873) preceded the decline of autocorrelation that mainly occurred between 1873 and 1969 (compare Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e with Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ee, \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ef \u0026amp; \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eg). The area and autocorrelation of grassland followed opposite trends (compare Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). Although the grassland area doubled between 1774 and 2022, its dispersed increase on soil groups where it was previously scarce reduced spatial autocorrelation. The autocorrelation of orchards was higher in 2022 than before, in spite of the area decline in the last time interval (compare Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eh). The contraction of this land-use to silt loam soil is explanatory for this finding (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh).\u003c/p\u003e \u003cp\u003eA complementary metric to evaluate the change of the landscape texture is the interspersion level, that refers to the spatial intermixing of different patch types, land-use classes in this case, without explicit reference to the dispersion level that we quantified by means of global Moran\u0026rsquo;s I. In 1774 and 1873, the interspersion level was less than 50% for approximately two thirds of northern Belgium (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e). River valleys, where several land-use classes converged, were accentuated by high interspersion levels, whereas vast areas of arable land, heathland \u0026amp; dune, and forest displayed the lowest interspersion levels (compare Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Interspersion increased between 1774 and 1873 in the Northeast, as a consequence of the conversion of heathland \u0026amp; dune to forest and arable land (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Between 1873 and 1969 the interspersion of land-use classes increased sharply (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e) and masked the hydrological structure (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), as a result of the dispersed increase of grassland and built-up area \u0026amp; garden (compare Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Expanding cities, e.g. Brussels, Antwerp, Ghent and Bruges, have become islands of homogeneous land-use with little land-use interspersion in 1969 and 2022 (compare Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eLand reclamation\u003c/h2\u003e \u003cp\u003eThe oldest map in our study, finished in 1774, depicts the Southern or Austrian Netherlands that belonged to the Habsburg Empire. The map is a snapshot of the landscape at the end of the Ancient R\u0026eacute;gime, before the socioeconomic changes inflicted by the French invasion in 1795 and industrialization transformed the landscape \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. It is assumed that agricultural expansion was the primary response to population growth at that time causing the cropland area in most European countries to peak around 1900 \u003csup\u003e19\u003c/sup\u003e. The land-use changes in northern Belgium in the first time interval (1774\u0026ndash;1873) support this assumption. As a result heathland and dune vegetation, inland marshes, long-established forests, and the intertidal zone, i.e. natural and semi-natural land-use, declined in the same time interval.\u003c/p\u003e \u003cp\u003eHeathland \u0026amp; dune vegetation and inland marshes, mostly located on sand soil, still covered more than 12% of northern Belgium in 1774. These were used as common land for pasturing, cutting of peat, digging of loam, and wood gathering. In Northwestern Europe the degradation of common land-use geared up in the last decades of the 18th century \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Within a century, the areas of heathland \u0026amp; dune, inland marshes, and the intertidal zone in northern Belgium were approximately reduced by half, and in the case of heathland \u0026amp; dune, the decline continued up to 2022. The shift of land-use on sand soil could indicate that forest plantation and reclamation to arable land respectively accounted for two thirds and one third of the heathland \u0026amp; dune area loss between 1774 and 1873. Our cartographic analysis also indicates that the area of inland marshes and the intertidal area was already reduced with 50% between 1774 and 1873, whereas the worldwide loss of wetlands is assumed to have largely taken place in the 20th century \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. This finding confirms that the reclamation of marshes, i.e. wetlands with a permanent high water table, could have preceded the decline of other wetlands, e.g. wet grasslands \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAlthough the increasing availability of fossil pit coal fueled the early industrialization of Belgium from 1750 onward \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, forests remained important for energy supply at least until the first decades of the 19th century \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. However, in the aftermath of the French Revolution, forests of the clergy and aristocracy were confiscated, privatized, and converted to farmland \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Furthermore, a famine caused by the potato blight (1846\u0026ndash;1850) exacerbated deforestation in Northwestern Belgium \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. In many parts of Europe forest cover was at a minimum in the 19th century as a result of agricultural expansion \u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. The already low total forest cover in northern Belgium did not change much between 1774 and 1873, as the conversion of heathland and marshes to plantation forests with coniferous trees \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e compensated for the loss of 50% of the forest area on fertile soils. Low net forest area changes thus concealed the erosion of long-established ancient forests with a high naturalness \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eAgricultural intensification\u003c/h2\u003e \u003cp\u003eEuropean agriculture was challenged by globalization after 1870, in particular by the cheap import of overseas cereals \u003csup\u003e28 29\u003c/sup\u003e. At that time the import and industrial production of fertilizers also stimulated agricultural intensification \u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Both facilitated a strong increase of livestock production at the end of the 19th century and even more in the 20th century \u003csup\u003e31 32 33\u003c/sup\u003e. In England \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e and the Netherlands \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e this caused a shift from cropland to pasture, and our study revealed that the grassland area doubled in northern Belgium between 1873 and 1969. By contrast, the increased focus on livestock production did not inflict such land-use change in Germany \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e, Spain \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e, Denmark \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e and Czech Republic \u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. As a result of intensification, Belgian agriculture not only shifted to livestock production, but also to horticulture and fruit production \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. The area of orchards for fruit production increased between 1873 and 1969, and became concentrated on silt loam soils in the Southeast of our study area whereas dispersed farm-based orchards disappeared \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eUrbanization\u003c/h3\u003e\n\u003cp\u003eThe built-up \u0026amp; garden area increased from 6.4% in 1774 to 29.3% in 2022, which is more or less proportional to the population growth in that time interval, from between 1,500,000 and 2,000,000 inhabitants in 1774 to 8,000,000 in 2022. Population growth exceeded the area increase between 1774 and 1969, and can be explained by urban transition during industrialization in the 19th and the first half of the 20th centuries \u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. If we only consider the last time period (1969\u0026ndash;2022), the area increase of built-up \u0026amp; garden (+\u0026thinsp;76%) far exceeded population growth (+\u0026thinsp;23%), which points to urban sprawl \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. Our area-wide results confirm previous assessments of urban sprawl based on indirect population data at the level of municipalities \u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e and case studies \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. Urban sprawl in northern Belgium often manifests itself as ribbon development guided by historical patterns of roads, dikes and villages and is facilitated by a lack of a policy to prevent this self-reinforcing process \u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. As a result, this region is a hotspot of urbanization in recent decades \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe increase of infrastructure, buildings and gardens after 1969 was proportional on all soil groups, so polder, wet \u0026amp; organic, and alluvial soils were not avoided. This is a confirmation of case studies of river valleys \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e and the assessment that new settlements established in Belgium between 1985 and 2015 are evenly distributed over flooding risk categories \u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003eLandscape transformation\u003c/h3\u003e\n\u003cp\u003eLandscape coherence, defined as the overall association between land-use and soil patterns \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e, first increased in northern Belgium between 1774 and 1873 as a result of land reclamation. In a similar way, by conversion of forest to arable land, the natural potential of fertile sites was further unleashed in 19th century Germany \u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e and Poland \u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. By contrast, socioeconomic drivers explained land-use changes between 1860 and 1958 in a mountainous French Mediterranean landscape more than biophysical drivers did, but this was reversed afterwards \u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAs opposed to these areas in Germany, Poland, and France, that are still rural, the association between site conditions and land-use progressively weakened in northern Belgium after 1873. A similar trend occurred in the Netherlands between 1900 and 1990, also with an increase of grassland on a wide range of soil groups favoring this trend and a concentration of forest on coarse sand soil opposing it \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. A study of regional differences at a single time slice in Estonia confirmed that land-use intensification increased the similarity between landscapes and thus the homogenization of the landscape structure \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe 3 drivers of land-use change, being land reclamation, agricultural intensification, and urbanization followed each other in time. The rise of the land-use interspersion level could indicate that landscape transformation culminated between 1873 and 1969. Up to 1873 land-use reflected the soil zonation and delineated distinct landscapes, e.g. grassland in valleys and polders, arable land on plateaus of sandy loam and silt loam, and also adjacent to villages on sand soils. These villages and surrounding arable lands on sand soil were separated by vast areas of heathland and marshes. From 1969 onwards, land-use is interspersed to a high degree, in particular by the grassland area that expanded beyond polders and valleys and by progressive urbanization that manifested itself as urban sprawl. As a result, northern Belgium transformed from a mostly rural region with distinct landscapes, to a homogenized peri-urban region. This trajectory could be illustrative for other peri-urban regions of Europe that show similar growth rates of the surface area (+\u0026thinsp;78%) and the population (+\u0026thinsp;33%) since the mid-1950\u0026rsquo;s \u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e "},{"header":"Methods","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003eStudy area and soil typology\u003c/h2\u003e \u003cp\u003eNorthern Belgium, composed of the administrative Flemish and Brussels capital regions, is a flat or undulating region with an altitude below 290 m above the North Sea level (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea), at the southwest side of the North European Plain (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Topsoils of the North European Plain mainly consist of varying fractions of periglacial pleistocene aeolian sand and silt deposits \u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. As a result there is a north-south zonation of sand soil, sandy loam soil, and silt loam soil, also present in northern Belgium (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). To determine the association between land-use and soils, the Belgian soil map was reclassified to 7 soil groups based on soil texture and drainage class that are robust soil properties determined by long-term pedogenic processes \u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. The Belgian soil map (1: 20,000) resulted from the National Soil Survey, carried out between 1947 and 1971. Sites classified as disturbed soil, built-up area, water, or not surveyed areas such as military zones, were grouped into a \u0026lsquo;not specified\u0026rsquo; soil group.\u003c/p\u003e \u003cp\u003eThe climate of northern Belgium is temperate oceanic, as indicated by averages calculated for 1991\u0026ndash;2020 \u003csup\u003e54\u003c/sup\u003e. The annual temperature was 11.0\u0026deg;C, and the average daily maximum and minimum temperatures equaled 14.7\u0026deg;C and 7.3\u0026deg;C, respectively. On 189.8 days with precipitation per year, an average total precipitation of 837.1 mm was measured. From the North Sea shore in the West towards the eastern border, there is a slight increase in continentality.\u003c/p\u003e \u003cp\u003eAt the end of the 18th century the population of northern Belgium was estimated between 1,500,000 and 2,000,000 inhabitants \u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e, increasing to 3,000,000 in 1873, and 6,500,000 in 1969. Presently there are 8,000,000 inhabitants \u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eHistorical and present-day maps\u003c/h2\u003e \u003cp\u003eWe used the maps of Ferraris that cover the Austrian Netherlands before the independence of Belgium, the first edition of topographical maps of Belgium by D\u0026eacute;p\u0026ocirc;t de la Guerre and the topographical maps of Belgium by the Military Geographical Institute (Supplementary Table S5) for the segmentation of historical land-use. These 3 historical maps have a native scale of 1:11,520, 1:20,000 and 1:25,000, respectively, and are available as high-quality tiled maps at the Geopunt and Cartesius geoportals. The present-day land-use map of northern Belgium is composed of the map of land-use in 2022 of the region of Flanders \u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e, complemented with a similar land-use map of the capital region of Brussels. The final year of the observations is used to refer to the maps, i.e. 1774, 1873, 1969, and 2022, respectively.\u003c/p\u003e \u003cp\u003eThe 1774 map does not completely cover the present-day territory of Belgium and not mapped communities along the present-day border (1.0% of the study area) are excluded from spatial analyses. Areas of the land-use 2022 map assigned to sports and recreation (2.9% of the study area) are also excluded, as they are not discerned on historical maps and presently consist of an unknown mixture of land-use, such as forest, grassland, water, buildings and infrastructure.\u003c/p\u003e \u003cp\u003eAlthough the georeferencing of the tiled historical maps is of high quality \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, positional errors of the order of 10s of meters are still likely. Positional errors are propagated in the overlay of time slices and influence landscapes indices calculated on it \u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. To avoid error propagation we did not analyze land-use conversions in a direct way, using an overlay of 4 time slices, but calculated proportions of land-use classes in soil groups and vice versa, to evaluate the shift of land-use through time in relation to soils.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eLand-use segmentation and validation\u003c/h2\u003e \u003cp\u003eWe used OrthoSeg \u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e, an open source GeoAI software package, for image segmentation applied to historical maps. The specific legend and drawing of each historical map series required a specific number of classes for segmentation, further called segmentation classes (Supplementary Table S5). The training was based on area-wide manual digitization of segmentation class polygons, in square boxes that measured 256 x 256 m for the 1774 and 1873 maps, and 128 x 128 m for the 1969 map. The prediction ran on downloaded 2,048 x 2,048 pixel map tiles, with a pixel resolution of 1 m for the 1774 and 1873 maps and 0.5 m for the 1969 map, and with an overlap of 128 pixels for the 1774 and 1873 maps, and 256 pixels for the 1969 map. A first run based on few and simple training data for the initially discerned segmentation classes was followed by a desktop evaluation of the outcome. Next additional training boxes were digitized to correct segmentation errors or remove gaps and thus improve the result of the following run. If necessary, additional segmentation classes were discerned and training data of the previous run were adjusted accordingly. This process was repeated until the desktop evaluation was favorable, and comprised 12, 13, and 6 iterations for the 1774, 1873, and 1969 maps, respectively (Supplementary Table S5).\u003c/p\u003e \u003cp\u003eThe output of the segmentation were polygon layers for each historical time slice, with 16 (1774), 19 (1873), and 29 (1969) segmentation classes (Supplementary Table S6). These segmentation classes were clustered to 9 common land-use classes and a class with not specified land-use (Supplementary Table S6), that also included areas of the 2022 map assigned to sports and recreation.\u003c/p\u003e \u003cp\u003eThe validation used area-based error matrices that take into account the selection bias introduced by the stratified sampling of the validation points, to calculate the estimated areas and confidence intervals (CI), and the producer\u0026rsquo;s, consumer\u0026rsquo;s and overall accuracies \u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e (Supplementary Tables S1-4).\u003c/p\u003e \u003cp\u003eFor the validation of the segmented historical land-use maps we selected 50\u0026ndash;100 points per segmentation class with randomly generated coordinates. The selected validation points were assigned to the discerned segmentation classes by visual interpretation of the historical map scans, in a QGIS environment. Next, the validation points were clustered to the 9 land-use classes in a similar way as the outcome of the segmentation (Supplementary Table S6).\u003c/p\u003e \u003cp\u003eWe used the 2021 European Land-Use/Land Cover Area Frame Statistical Survey (LUCAS) point data \u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e to validate the thematic accuracy of the 2022 land-use map. The LUCAS validation points are assigned to harmonized land-use and land cover classes by means of a field survey, following a stratified random sampling on a 2 x 2 km grid covering Europe. The land-use class \u0026lsquo;heathland \u0026amp; dune\u0026rsquo; in our study is assigned to shrubland in the LUCAS typology, whereas shrubland on historical maps was mostly assigned to forest based on the common green legend coloration. Therefore we did not include survey points of the LUCAS land cover categories D (Shrubland) and F (Bare soil and Lichens) (50 points) in the validation. Consequently, the area of land-use class heathland \u0026amp; dune was not validated or adjusted based on the area-based error matrix (Supplementary Table S4). The LUCAS category of \u0026lsquo;temporary grasslands\u0026rsquo; (B55) included 180 validation points divided over our land-use classes \u0026lsquo;arable land\u0026rsquo; (89 points) and \u0026lsquo;grassland\u0026rsquo; (91 points). Both assessments were considered correct, as nowadays in our study area there is no clear distinction between temporary grassland, classified to LUCAS land cover category B (Cropland) and intensively used and fertilized permanent grassland, classified to LUCAS category E (Grassland).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eSpatial analyses\u003c/h2\u003e \u003cp\u003eWe converted the polygon layers of historical land-use to grids with the same extent and 10 x 10 m cell size as the land-use 2022 map for spatial analyses on land-use classes: the association of land-use classes with soil groups, the spatial autocorrelation of land-use classes, and the interspersion of land-use classes.\u003c/p\u003e \u003cp\u003eWe analyzed the spatial association of 9 land-use classes with 7 soil groups at 4 time slices and used for this purpose the global V-measure, and its decomposition into the homogeneity and completeness measures \u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e, of land-use classes in soil groups and vice versa. Calculations were performed in Python using the GDAL and Scikit-Learn packages.\u003c/p\u003e \u003cp\u003eThe four land-use grid layers were also used to evaluate the spatial autocorrelation of land-use classes, i.e., the degree of fragmentation or clustering of land-use. For this purpose we converted the grid layers to binary maps, for each land-use class at each time slice, and calculated the corresponding Moran\u0026rsquo;s I values \u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e in Python using the GDAL, Numpy and SciPy packages. Incremental spatial lags, ranging between 10 m and 10,000 m, were used to assess spatial autocorrelation over various distances.\u003c/p\u003e \u003cp\u003eThe interspersion was calculated with the r.neighbors package of grass in QGIS on a focal area with 500 m radius. The interspersion of a cell is the percentage of cells containing values which differ from the values assigned to the center cell, in the circular neighborhood with a 500 m radius, plus 1.\u003c/p\u003e \u003c/div\u003e "},{"header":"Declarations","content":"\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003eData availability\u003c/h2\u003e\n \u003cp\u003eThe tiled historical maps of 1774 can be viewed and accessed at the Geopunt regional geodata portal of Flanders: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.vlaanderen.be/datavindplaats/catalogus/ferraris-kaart-kabinetskaart-der-oostenrijkse-nederlanden-en-het-prinsbisdom-luik-1771-1778\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eTiled historical topographical maps are online available at the federal Cartesius geodata portal:\u003c/p\u003e\n \u003cp\u003eTopographical map of 1873: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://wmts.ngi.be/arcgis/rest/services/seamless_carto__default__3857__140/MapServer/tile/{z}/{y}/{x\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e}\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eTopographical map of 1969: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://wmts.ngi.be/arcgis/rest/services/seamless_carto__default__3857__1100/MapServer/tile/{z}/{y}/{x\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e}\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe maps of segmented historical land-use can be viewed and accessed at: \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003ettps://\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.vlaanderen.be/datavindplaats/catalogus/digitalisatie-historisch-landgebruik-en-landgebruiksveranderingen-in-vlaanderen-1778-2022\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003eCode availability\u003c/h2\u003e\n \u003cp\u003eThe applied OrthoSeg open source software package is developed for the segmentation of remote sensing images and maps shared by geoportals \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e and can be accessed at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/orthoseg/orthoseg/wiki\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eConceptualization: L.D.K., P.R., L.P. Methodology: P.R., L.D.K., L.P., F.P., S.T., T.P. Visualization: L.D.K., F.P. Funding acquisition: J.V.V. Writing: L.D.K., F.P., L.P., J.V.V., P.R.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLe Provost G et al (2020) Land-use history impacts functional diversity across multiple trophic groups. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 117, 1573\u0026ndash;1579\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFoster D et al (2003) The Importance of Land-Use Legacies to Ecology and Conservation. Bioscience 53:77\u0026ndash;88\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchulp CJE, Levers C, Kuemmerle T, Tieskens KF, Verburg PH (2019) Mapping and modelling past and future land use change in Europe\u0026rsquo;s cultural landscapes. Land Use Policy 80:332\u0026ndash;344\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchirpke U et al (2023) Past and future impacts of land-use changes on ecosystem services in Austria. J Environ Manage 345:118728\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eB\u0026uuml;rgi M, \u0026Ouml;stlund L, Mladenoff DJ (2017) Legacy Effects of Human Land Use: Ecosystems as Time-Lagged Systems. Ecosystems 20:94\u0026ndash;103\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWinkler K, Fuchs R, Rounsevell M, Herold M (2021) Global land use changes are four times greater than previously estimated. Nat Commun 12:2501\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui Q et al (2014) Historical land-use and landscape change in southern Sweden and implications for present and future biodiversity. Ecol Evol 4:3555\u0026ndash;3570\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u003cem\u003eEarth Syst. Sci. Data\u003c/em\u003e 15, 1005\u0026ndash;1035 (2023)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaprotny D, Mengel M (2023) Population, land use and economic exposure estimates for Europe at 100 m resolution from 1870 to 2020. Sci Data 10:372\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlah B (2009) Historical maps and their application in landscape ecological research. Ekologia 28:143\u0026ndash;151\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoran C, Haegi S, Ginzler C (2018) Comparing historical and contemporary maps - a methodological framework for a cartographic map comparison applied to Swiss maps. Int J Geogr Inf Sci 32:2123\u0026ndash;2139\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCousins SAO, Auffret AG, Lindgren J, Tr\u0026auml;nk L (2015) Regional-scale land-cover change during the 20th century and its consequences for biodiversity. Ambio 44:17\u0026ndash;27\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePicuno P, Cillis G, Statuto D (2019) Investigating the time evolution of a rural landscape: How historical maps may provide environmental information when processed using a GIS. Ecol Eng 139:105580\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLathouwers E, Segers Y, Verstraeten G (2023) Reconstructing valley landscapes. GIS-analyses of past land use changes in three Flemish river valleys since the late 18th century. Land Use Policy 135:106960\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang Y, Gao S, Roth RE (2024) Artificial intelligence studies in cartography: a review and synthesis of methods, applications, and ethics. Cartogr Geogr Inf Sci 51:599\u0026ndash;630\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u0026Aring;gren AM, Lin Y (2024) A fully automated model for land use classification from historical maps using machine learning. Remote Sens Appl Soc Environ 36:101349\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVervust S, Bo\u0026ugrave;\u0026ugrave;aert MC, De Baets B, Van de Weghe N, De Maeyer P (2018) A Study of the Local Geometric Accuracy of Count de Ferraris\u0026rsquo;s Carte de cabinet (1770s) Using Differential Distortion Analysis. Cartogr J 55:16\u0026ndash;35\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaeyer PD (2000) Base maps in Belgium. Belg Rev Belge G\u0026eacute;ographie 165\u0026ndash;172. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4000/belgeo.13977\u003c/span\u003e\u003cspan address=\"10.4000/belgeo.13977\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYe Y et al (2023) Reconstruction of cropland change in European countries using integrated multisource data since AD 1800. Boreas 52:60\u0026ndash;77\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Moor M, Shaw-Taylor L, Warde P (2002) Comparing the historical commons of north west Europe. An introduction. The management of common land in north west Europe, c. 1500\u0026ndash;1850, vol 8. Brepols, pp 15\u0026ndash;31\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFluet-Chouinard E et al (2023) Extensive global wetland loss over the past three centuries. Nature 614:281\u0026ndash;286\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGodet L, Thomas A (2013) Three centuries of land cover changes in the largest French Atlantic wetland provide new insights for wetland conservation. Appl Geogr 42:133\u0026ndash;139\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRyckbosch W, Saelens W (2023) Fuelling the urban economy: A comparative study of energy in the Low Countries, 1600\u0026ndash;1850. Econ Hist Rev 76:221\u0026ndash;256\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTallier P-A, Verboven H, Vandekerkhove K, Baet\u0026eacute; H, Verheyen K (2018) CHAPTER 3 State Forestry in Belgium since the End of the Eighteenth Century. In: Oosthoek KJ, H\u0026ouml;lzl R (eds) Managing Northern Europe\u0026rsquo;s Forests: Histories from the Age of Improvement to the Age of Ecology. Berghahn Books, pp 92\u0026ndash;129. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1515/9781785336010-007\u003c/span\u003e\u003cspan address=\"10.1515/9781785336010-007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerheyen K, Bossuyt B, Hermy M, Tack G (1999) The land use history (1278\u0026ndash;1990) of a mixed hardwood forest in western Belgium and its relationship with chemical soil characteristics. J Biogeogr 26:1115\u0026ndash;1128\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaplan JO, Krumhardt KM, Zimmermann N (2009) The prehistoric and preindustrial deforestation of Europe. Quat Sci Rev 28:3016\u0026ndash;3034\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerg\u0026egrave;s L, Dupouey J-L (2021) Historical ecology and ancient forests: Progress, conservation issues and scientific prospects, with some examples from the French case. J Veg Sci 32:e12846\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO\u0026rsquo;Rourke KH (1997) The European Grain Invasion, 1870\u0026ndash;1913. J Econ Hist 57:775\u0026ndash;801\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Dijck M, Truyts T (2014) The Agricultural Invasion and the Political Economy of Agricultural Trade Policy in Belgium, 1875\u0026ndash;1900. SSRN Scholarly Paper at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2139/ssrn.2378740\u003c/span\u003e\u003cspan address=\"10.2139/ssrn.2378740\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrotmann C, Herment L, Page A (2021) Fertilisers in the Long 19th Century and Beyond: Usage, Commercialisation and Production (c 1800\u0026ndash;1939). \u003cem\u003eJahrb. F\u0026uuml;r Wirtsch. Econ. Hist. Yearb.\u003c/em\u003e 62, 1\u0026ndash;18\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFederico G, THE GROWTH OF WORLD AGRICULTURAL, PRODUCTION (1938) 1800\u0026ndash; in \u003cem\u003eResearch in Economic History\u003c/em\u003e vol. 22 125\u0026ndash;181 (Emerald Group Publishing Limited, 2004)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSwinnen JFM (2010) Agricultural Protection Growth in Europe, 1870\u0026ndash;1969. In: Anderson K (ed) The Political Economy of Agricultural Price Distortions. Cambridge University Press, Cambridge, pp 141\u0026ndash;161. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1017/CBO9780511778964.008\u003c/span\u003e\u003cspan address=\"10.1017/CBO9780511778964.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamankutty N et al (2018) Trends in Global Agricultural Land Use: Implications for Environmental Health and Food Security. Annu Rev Plant Biol 69:789\u0026ndash;815\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHowkins A (2021) Reshaping Rural England: A Social History 1850\u0026ndash;1925. Routledge, London. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4324/9780203421185\u003c/span\u003e\u003cspan address=\"10.4324/9780203421185\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBakker MM, Sonneveld MPW, Brookhuis B, Kuhlman T (2013) Trends in soil\u0026ndash;land-use relationships in the Netherlands between 1900 and 1990. Agric Ecosyst Environ 181:134\u0026ndash;143\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNiedertscheider M, Kuemmerle T, M\u0026uuml;ller D, Erb K-H (2014) Exploring the effects of drastic institutional and socio-economic changes on land system dynamics in Germany between 1883 and 2007. Glob Environ Change 28:98\u0026ndash;108\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Gonz\u0026aacute;lez M et al (2020) Agricultural Output: From Crop Specialization to Livestocking, 1900\u0026ndash;2008. In: de Gonz\u0026aacute;lez M et al (eds) The Social Metabolism of Spanish Agriculture, 1900\u0026ndash;2008: The Mediterranean Way Towards Industrialization. Springer International Publishing, Cham, pp 29\u0026ndash;68. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-030-20900-1_2\u003c/span\u003e\u003cspan address=\"10.1007/978-3-030-20900-1_2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLohrum N, Normand S, Dalgaard T, Graversgaard M (2024) Unveiling the frontiers: Historical expansion and modern implications of agricultural land use in Denmark. J Environ Manage 359:120934\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoravcova J, Moravcova V, Pavlicek T, Novakova N (2022) Land Use Has Changed through the Last 200 Years in Various Production Areas of South Bohemia. Land 11:1619\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBocquier P, Costa R (2015) Which transition comes first? Urban and demographic transitions in Belgium and Sweden. Demogr Res 33:1297\u0026ndash;1332\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnaap G-J, Song Y, Nedovic-Budic Z (2007) Measuring Patterns of Urban Development: New Intelligence for the War on Sprawl. Local Environ 12:239\u0026ndash;257\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMustafa A, Teller J (2020) Self-Reinforcing Processes Governing Urban Sprawl in Belgium: Evidence over Six Decades. Sustainability 12:4097\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePoelmans L, Van Rompaey A (2009) Detecting and modelling spatial patterns of urban sprawl in highly fragmented areas: A case study in the Flanders\u0026ndash;Brussels region. Landsc Urban Plan 93:10\u0026ndash;19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerbeek T, Boussauw K, Pisman A (2014) Presence and trends of linear sprawl: Explaining ribbon development in the north of Belgium. Landsc Urban Plan 128:48\u0026ndash;59\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuemmerle T et al (2016) Hotspots of land use change in Europe. Environ Res Lett 11:064020\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRapid Urban Growth in Flood Zones: Global Evidence since 1985 (2022) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1596/1813-9450-1014\u003c/span\u003e\u003cspan address=\"10.1596/1813-9450-1014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMander \u0026Uuml;, Uuemaa E, Roosaare J, Aunap R, Antrop M (2010) Coherence and fragmentation of landscape patterns as characterized by correlograms: A case study of Estonia. Landsc Urban Plan 94:31\u0026ndash;37\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWulf M, Jahn U, Meier K (2016) Land cover composition determinants in the Uckermark (NE Germany) over a 220-year period. Reg Environ Change 16:1793\u0026ndash;1805\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZgłobicki W, Gawrysiak L, Baran-Zgłobicka B, Telecka M (2016) Long-term forest cover changes, within an agricultural region, in relation to environmental variables, Lubelskie province, Eastern Poland. Environ Earth Sci 75:1373\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbadie J et al (2018) Forest recovery since 1860 in a Mediterranean region: drivers and implications for land use and land cover spatial distribution. Landsc Ecol 33:289\u0026ndash;305\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUrban sprawl in Europe - The ignored challenge. Eur Environ Agency \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.eea.europa.eu/publications/eea_report_2006_10/eea_report_10_2006.pdf\u003c/span\u003e\u003cspan address=\"https://www.eea.europa.eu/publications/eea_report_2006_10/eea_report_10_2006.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBertran P et al (2021) Revised map of European aeolian deposits derived from soil texture data. Quat Sci Rev 266:107085\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Keersmaeker L et al (2013) Application of the Ancient Forest Concept to Potential Natural Vegetation Mapping in Flanders, A Strongly Altered Landscape in Northern Belgium. Folia Geobot 48:137\u0026ndash;162\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKMI - Jaar KMI \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.meteo.be/nl/klimaat/klimaat-van-belgie/klimatologisch-overzicht/2023/jaar\u003c/span\u003e\u003cspan address=\"https://www.meteo.be/nl/klimaat/klimaat-van-belgie/klimatologisch-overzicht/2023/jaar\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlep PM (1991) Population Estimates of Belgium, by Province (1375\u0026ndash;1831). Soci\u0026eacute;t\u0026eacute; Belge de D\u0026eacute;mographie, Louvain-la-Neuve\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKey Figs (2023) | Statbel. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://statbel.fgov.be/en/news/key-figures-2023\u003c/span\u003e\u003cspan address=\"https://statbel.fgov.be/en/news/key-figures-2023\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePoelmans L, Janssen L, Hambsch L (2023) Landgebruik en ruimtebeslag in Vlaanderen, toestand 2022 - Onderzoeksportaal. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.friscris.be/nl/publications/landgebruik-en-ruimtebeslag-in-vlaanderen-toestand-2022(ab4c67af-a12e-4acf-9346-39df99589565).html\u003c/span\u003e\u003cspan address=\"https://www.friscris.be/nl/publications/landgebruik-en-ruimtebeslag-in-vlaanderen-toestand-2022(ab4c67af-a12e-4acf-9346-39df99589565).html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurnicki AC (2012) Impact of error on landscape pattern analyses performed on land-cover change maps. Landsc Ecol 27:713\u0026ndash;729\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoggemans P (2024) orthoseg. Zenodo \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5281/ZENODO.10340584\u003c/span\u003e\u003cspan address=\"10.5281/ZENODO.10340584\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlofsson P, Foody GM, Stehman SV, Woodcock CE (2013) Making better use of accuracy data in land change studies: Estimating accuracy and area and quantifying uncertainty using stratified estimation. Remote Sens Environ 129:122\u0026ndash;131\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThe thematic accuracy of Corine land cover 2000 - Assessment using LUCAS. Eur Environ Agency \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.eea.europa.eu/publications/technical_report_2006_7\u003c/span\u003e\u003cspan address=\"https://www.eea.europa.eu/publications/technical_report_2006_7\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNowosad J, Stepinski TF (2018) Spatial association between regionalizations using the information-theoretical V -measure\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoran PAP (1950) Notes on Continuous Stochastic Phenomena. Biometrika 37:17\u0026ndash;23\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLevin G, Groom G, Svenningsen SR (2024) Assessing long-term landscape dynamics based on automated production of land category layers from Danish late 19th century topographic maps. Preprint at. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21203/rs.3.rs-4021413/v1\u003c/span\u003e\u003cspan address=\"10.21203/rs.3.rs-4021413/v1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eM\u0026auml;yr\u0026auml; J, Kivinen S, Keski-Saari S, Poikolainen L, Kumpula T (2023) Utilizing historical maps in identification of long-term land use and land cover changes. Ambio 52:1777\u0026ndash;1792\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartinez T et al (2023) Deep learning ancient map segmentation to assess historical landscape changes. J Maps 19:2225071\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuggitt AJ et al (2023) Linking climate warming and land conversion to species\u0026rsquo; range changes across Great Britain. Nat Commun 14:6759\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO\u0026rsquo;Hara R, Marwaha R, Zimmermann J, Saunders M, Green S (2024) Unleashing the power of old maps: Extracting symbology from nineteenth century maps using convolutional neural networks to quantify modern land use on historic wetlands. Ecol Indic 158:111363\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSt\u0026aring;hl N, Weimann L (2022) Identifying wetland areas in historical maps using deep convolutional neural networks. Ecol Inf 68:101557\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEuro Dem | Eurogeographics \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mapsforeurope.org/datasets/euro-dem\u003c/span\u003e\u003cspan address=\"https://www.mapsforeurope.org/datasets/euro-dem\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5536645/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5536645/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWe quantified historical land-use with deep learning segmentation, applied to tiled historical maps, and identified 3 successive drivers of long-term (1774\u0026ndash;2022) landscape transformation in northern Belgium (13,800 km\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e). Between 1774 and 1873, land reclamation halved the area of natural and semi-natural land-use. Agricultural intensification was the main driver in the next time interval (1873\u0026ndash;1969), as the area of grassland and orchard doubled at the expense of arable land. Urbanization marked the last time interval (1969\u0026ndash;2022) and reduced agricultural land-use. The reclamation of fertile soils for agriculture and the shift of forests to sand soils previously covered by heathland first increased the association of land-use classes to soil groups. After 1873 this association progressively weakened by expansion of grasslands beyond valleys and polders and urbanization disregarding soils. A sharp rise of land-use interspersion indicated that landscape transformation culminated between 1873 and 1969 and resulted in the homogenization of previously distinct landscapes.\u003c/p\u003e","manuscriptTitle":"Two and a half centuries of land reclamation, intensification, and urbanization homogenized northern Belgium landscapes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-10 09:22:14","doi":"10.21203/rs.3.rs-5536645/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ced92e0e-3108-481b-8e12-b78e19d6348f","owner":[],"postedDate":"January 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":42648684,"name":"Scientific community and society/Geography"},{"id":42648685,"name":"Earth and environmental sciences/Environmental social sciences"}],"tags":[],"updatedAt":"2026-02-21T08:11:11+00:00","versionOfRecord":{"articleIdentity":"rs-5536645","link":"https://doi.org/10.1038/s41467-026-68594-y","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2026-01-21 05:00:00","publishedOnDateReadable":"January 21st, 2026"},"versionCreatedAt":"2025-01-10 09:22:14","video":"","vorDoi":"10.1038/s41467-026-68594-y","vorDoiUrl":"https://doi.org/10.1038/s41467-026-68594-y","workflowStages":[]},"version":"v1","identity":"rs-5536645","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5536645","identity":"rs-5536645","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}