Materials for Afforestation Reconstruction in Nowy Tomyśl Plain in the 16th Century

Tomasz Panecki, Tomasz Związek, The Tadeusz Manteuffel Institute of History of the Polish Academy of Sciences
technical editors A. Borek, G. Myrda

Introduction

Current works of the DHA related to the “Historical Atlas of Poland. Detailed maps of the 2^nd Half of the 16^th century” were the inspiration behind the afforestation reconstruction project (Słoń 2014). The methodology of the settlement network, administrative and ecclesiastical boundaries as well as natural landscape reconstruction for the early modern period requires using a large array of written and cartographic sources. Written sources come mostly from the second half of the 16^th c. and are used to reconstruct political and historical landscape (settlements and boundaries). Settlements are identified and localized on 20^th c. maps. Furthermore, their name, size, ownership, administrative and ecclesiastical affiliation are also determined. The elements of natural environment, such as hydrography and afforestation are obtained directly from the maps from the turn of 18^th/19^th c. and sometimes adjusted to 16^th c. state with the use of retrogressive method (Słoń 2014). Because of this methodology, there is a temporal inconsistency on the map: settlements and boundaries are 16^th c. depiction while physiography is almost 300 years newer. The originators of the series (almost 70 years ago) were aware of the problem, however due to the absence of cartographic sources for the 16^th c, it was necessary to use maps developed as early as possible, which cartometrcity and credibility were sufficient, i.e. maps from the turn of 18^th/19^th c. The result of this approach includes above-mentioned temporal mismatch which is illustrated on by the emptiness in Nowy Tomyśl Plain (figure 2). Depicted area lacks regular settlements, which were absent in 16^th c. and appeared at the beginning of 18^th c. as well as afforestation, which due to 18^th c. grubbing is not preserved on sources from the turn of the 18^th/19^th c. Although, the area was being explored in 16^th c.by the inhabitants of neighbouring villages, only periodical settlements were being established due to the acquisition of wood, charcoal and bog iron.

Fig. 1: The emptiness in the vicinities of Tomyśl in 16^th c., where Hollander colonization appeared in 18^th c. (See map)

Research hypotheses

The area of the analysis was in the 16^th c. certainly covered by forest, what is confirmed by written sources from this, and earlier period (see: "State Archive in Poznań" in literature references). From the beginning of 18^th c. to the early years of the 19^th c. the vicinities of Tomyśl city became a place of an intense so-called Dutch colonization, which had a great impact on the natural landscape (Rusiński 1947). The colonists, settling mostly on wetlands and forested areas, cleared the surrounding woods thereby changing relatively hostile to humanity areas into the place of constant residence. Due to the nature of location, as well as legal and ownership relations, the villages mostly took the form of scattered farm villages, where each colonist received a ground for his farm and arable land (Rusiński 1947).

There has been made an assumption that area of each farm’s arable land (acreage allocated to each colonist) should match with the area of cleared forest. The estimation of this value was possible through the use of an array of sources: written, cartographic and archaeological supplemented by the literature. In addition, in order to reconstruct the afforestation in a complex way, the aim of the study included the determination of forest type in the 16^th c. This kind of information is available through either historical and contemporary sources, such as: soil maps, geological maps and map of potential natural vegetation of Poland (Matuszkiewicz et al. 1995).

Data sources and thematic layers

Analyses were performed in GIS software (ArcGIS 10.2), which served as a platform for data integration. The first step in the research was to develop a spatial database with topographic data of study area, which should correspond chronologically to the turn of 18^th/19^th c. For this purpose, old maps (e.g. the Gilly-Cron map, the Urmesstischblätter map as well as the Messtischblätter map), manuscripts with the privileges given the settlers by land owners (see: "State Archive in Poznań" in literature references) along with the information from the literature (Rusiński 1947) was used.

The structure and scope of thematic layers that has been developed is presented in Table 1. Layers containing settlements, administrative boundaries, natural landscape, archaeological data and the data for the afforestation reconstruction are covered. Vector layers were supplemented by raster layers: georeferenced and mosaicked old maps covering the study area followed by contemporary topographic maps served through the WMS service from the State Geoportal. Individual layers have been assigned to parent categories, which allow easier navigation through the site. This structure below also contains a file name in *.shp format for each layer because of the possibility of direct file download through WebGIS application.

ID	Layer Group	Layer	File Name	Layer Type
1	Settlements	Historical Atlas of Poland - settlements	HAP_settle_16c.shp	Vector/External Data
2	Settlements	Hollander Project - settlements	Holl_settle_19c.shp	Vector
3	Administrative Units	Historical Atlas of Poland - administrative units	HAP_adm_16c.shp	Vector/External Data
4	Administrative Units	Historical Atlas of Poland - ecclesiastical units	HAP_ecc_16c.shp	Vector/External Data
5	Administrative Units	Hollander Project - administrative units	Holl_adm_19c.shp	Vector
6	Natural Landcape	Urmesstischblatter - forests	UMSB_forests.shp	Vector
7	Natural Landcape	Urmesstischblatter - rivers	UMSB_hydro_line.shp	Vector
8	Natural Landcape	Urmesstischblatter - lakes	UMSB_hydro_polygon.shp	Vector
9	Natural Landcape	Gilly-Cron - forests	GK_forests.shp	Vector
10	Archeology	AZP	AZP.shp	Vector
11	Archeology	AZP index (WMS)	N/A	OGC service
12	Forest reconstruction data	10 hectares Buffer	Buffer_10ha.shp	Vector
13	Forest reconstruction data	15 hectares Buffer	Buffer_15ha.shp	Vector
14	Forest reconstruction data	20 hectares Buffer	Buffer_20ha.shp	Vector
15	Forest reconstruction data	25 hectares Buffer	Buffer_25ha.shp	Vector
16	Forest reconstruction data	Thiessen polygons	Thiessen_polygon.shp	Vector
17	Forest reconstruction data	Vegetation type	Vegetation_type.shp	Vector
18	Forest reconstruction data	Hollander Project - afforestation reconstruction	Holl_forests_16c.shp	Vector
19	Raster Basemaps	Gilly-Cron Map, [GK], 1:50 000, 1793	N/A	Raster
20	Raster Basemaps	Urmesstichblatter Map, [UMSB], 1:25 000, 1828-1832	N/A	Raster
21	Raster Basemaps	Messtischblatter Map [MSB], 1:25 000, 1839-1940	N/A	Raster
22	Reference Data	Topographic Map of Poland (WMS)	N/A	OGC service

Table 1: Data structure in WebGIS application

Raster layers

Old maps were the basis for vector layers development. The earliest map used for the research was the manuscript map Karte von Südpreußen by David Gilly and Cron elaborated in 1793. At 89 sheets developed in the scale approx. 1:50 000 the area occupied by Prussia after the second partition of Poland (1793) is shown. Map, despite its large scale is not precise due to applied mathematical basis – mapping was not preceded by triangulation and only compass was used. The biggest advantage of this map however, is the fact that it presents the reality from the end of the 18^th c., i.e. before the landscape and administrative changes that took place in the early 19^th c. (Scharfe 1972 Medyńska-Gulij & Lorek 2008). Three sheets were used for the analysis (Zbąszyń, Buk and Kamienna). To avoid overlapping sheets’ margins, they were cut to the frame and mosaicked in graphics editor. Map prepared in this way was georeferenced ArcGIS 10.2 using linear transformation and 4 control points. This map was the source for afforestation layer from the end of 18^th c. ("Gilly-Cron - forests").

Another map used in the research was elaborated by the Prussians in the beginning of 19^th c. and was called the Urmesstischblätter (scale: 1:25 000). The plan to develop a large-scale topographic map of Prussia was established in 1816. The military carried out the triangulation of the Prussian state, which allowed to determine the precise geodetic network. The officers of the Prussian army conducted field measurements including on the map almost all elements of the landscape thanks to its large scale and minor cartographic generalization. Although theoretically elaborated using single and consistent instruction, individual topographers added to their sheets new elements in the way, that the individual sheets are in this respect somewhat unique (Lorek 2011). Map was not intended for publication and remained as manuscript, and on its basis maps in smaller scale were to be published (Scharfe 1972 Konias 2010).

The last map which was analysed was the successor of the Urmesstischblätter. Elaborated in the late 19^th c. also in the scale of 1:25 000, the Messtischblätter inherits primarily size of each sheet in arc units (6' φ for 10' λ). Map elaboration was preceded by a completely new triangulation and was also – in contrast to the previous version – based on a fixed a catalogue of symbols (Konias 2010).

Georeferencing in those two cases was different then in Gilly-Cron map. Due to the same angular size of the sheets (Urmesstischblätter and Messtischblätter), a spatial index was created in ArcGIS having proper dimensions and datum (Deutsches Hauptdreiecksnetz). The index served as the basis for georeferencing and enabled to clip sheets’ frames and map mosaicking (Panecki 2014). Any possible shortcomings of georeferenced were corrected afterwards. In both maps, the position errors do not exceed 200 meters.

Vector layers

Vector layers are divided into two categories: developed during the project and external. The outer layers contain spatial data from already-mentioned database of Polish Territories of the Crown in the 16^th c. available on AtlasFontium.pl. Within this data settlements, administrative and ecclesiastical boundaries are included (Polish Territories ... 2016).

Other layers are the result of the project. The so-called Dutch settlements, which can be the basis for afforestation reconstruction are in the vast majority scattered farm villages, so we decided to take each single farm as a reference, which are in “point” topology in the database (layer "Hollander Project - settlements"). Some problems occurred during the data acquisition and were related to the specificity of morphological type of settlements, as not all of them consisted from scattered farms. The village "Borujskie Stare Olędry" is a good example: there was a church square in the centre ("Kirchpltaz" on Urmesstischblätter) consisting of built-up areas and a number of individual buildings (figure 2). In this case, we decided to take the 3 buildings as one point (to model exactly one farm), as a single farm consisted mostly of 3 houses. This simplification was caused by the desire to keep a model depiction of the phenomenon, although in the course of future works on the project, we aim to change the topology of settlements to polygon (built-up area, farm) and point (single buildings).

Fig. 2: Kirchpltaz along with vectorised Hollander farms (See map)

The next step was the reconstruction of villages’ boundaries, as the historical data from written sources was referenced to each village. Vectorised farms, Urmesstischblätter’s administrative borders depiction and names of villages where the basis for boundaries reconstruction. It was not possible, however, to develop a complete reconstruction on this basis due to an incomplete drawing of borders on the map and ambiguous spatial reference of labels. Adding more information to the analysis was possible by involving later Messtischblätter where almost every group of farms was supplemented with a label: "zu [name of the village]", what allowed – on the basis of the retrogressive method – assigning to each farm its village identifier and its name in the consequence. The prepared data served as the basis for boundaries drawing through modelling by Thiessen polygon (layer "Thiessen polygons") generated around each farm (each point). Furthermore, polygons were dissolved by the attribute "name", and after cartographic refinement (including matching borders to natural barriers) they made as villages’ boundaries (Szady 2010). Finally, attribute information was joined to this layer and contains data about the date and type of location area, population and quantity of houses (“Hollander Project - administrative units”).

Fig. 3: Detailed labels of farms’ affiliation to particular village on Messtischblätter map (See map)

Archaeological data is another data type. Layer "AZP" (“Archeologiczne Zdjęcie Polski”, English: “Archeological Survey of Poland”) mainly contains information about the ceramic discovered on the surface during field works conducted since the 70s of the 20^th c. (Woyda 1975). There is also information concerning tile fragments as well as remains of 17^th c. glassworks on the studied area. From this collection we excluded points, for which the number of ceramic pieces for one position of AZP was less than 5, considering them to be unreliable in the research. Moreover, interpreting this set of data is not straightforward, because in the case of the Nowy Tomyśl Plain, archaeologists did not realize the historical past of the region until the end of the survey – this seems to be a reason of identifying more than 100 points defined as settlements, when in fact there were over 20 scattered so-called Dutch colonies. It is also incomprehensible why the area near the medieval-origin blacksmith located east of the village Boruja was not surveyed. Layer "AZP" was also limited by us to the materials dating back to the early (EM) and late Middle Ages (LM), and the modern era defined as the period from the mid/late 16^th to the early 19^th c. (M). This layer was supplemented with WMS with the sheet sections of AZP (layer: "AZP index (WMS)").

Reconstruction of afforestation cannot take place without information about the natural landscape. Urmesstischblätter (approx. 1830) was the source for hydrography (layer "Urmesstischblatter - rivers" and "Urmesstischblatter - lakes") and afforestation (layer "Urmesstischblatter - forests"). Data with afforestation in approx.. 1830 was supplemented by nearly 40 years earlier (Gilly-Cron), which was intended to observe trends in land cover change (layer "Gilly-Cron - forests").

The service also includes layers directly related to the afforestation reconstruction, such as square buffers in 10, 15, 20 and 25 hectares generated around each farm, which are to symbolize the shape of farmland (the colonists most often received grounds in the form of a field close to the square), vegetation types derived from old and contemporary maps, and layer containing the reconstructed forest in 16^th c.

Methods and outcomes

After obtaining information from sources and spatial database development, the next step of the study was to estimate the area of each farm along with its arable land, which should correspond to the area of the cleared forest. For this purpose, different approaches were used specific to geography, history and archaeology. Applying various methods helped to verify the results obtained by different approaches. Almost all calculations are aggregated to polygon layer of settlements’’ boundaries (“Hollander Project - administrative units”).

Analysis of sources and historical literature has shown that the average land acreage given to settlers is approx. 16-24 hectares, although farms with less land were also possible (Rusiński 1947). Colonists’ arable land usually had a form similar to square and the farm containing buildings (most commonly 3) was located in the centre of it (Pelczyk 1996). Therefore, to model the acreage of each land, a point topology was adopted (one point – one farm) as well as 10, 15, 20 and 25 hectares square-shaped buffers around them (layer "Buffer_XXha"). Unfortunately it is not possible to calculate exact acreage of farms’ arable land, because we have data only from the moment of their establishment. These values may have changed between this date (mostly approx. 1750), and the time of cartographic depiction which we have (18^th/19^th c.). Moreover, the units of measure used in those times are somewhat uncertain – unit “włóka” used most commonly by the settlers may vary from 7 to 18 hectares.

Another possibility to estimate the arable land acreage is strictly geographical method. Its determination should be possible by calculating the area of non-forest area within each village and dividing this value by the number of farms within the village. Results of this calculation are in the range of from 7.4 to 15.6 hectares of forest per farm with an average approx. 11.3 hectares, and standard deviation of 2.4. They were also some "model" villages for which these calculations may be the most reliable, i.e. villages located in glades (figure). For them, the average deforested area by a single farm is approx. 10.3 hectares.

Fig. 4: Villages located on glades can be a model example of deforested area (See map)

Another method, which can be called geostatistical is based on creating Thiessen polygons around all farms (layer "Thiessen polygon"). The area of each polygon whose boundaries are exactly half the distance between adjacent points should correspond with the total forest area attributable to a single farm before its location. The results obtained by this method are quite diverse ranging from 7 to 35 hectares with an average of approx. 13.8 hectares and 10 hectares dominant. The standard deviation of the set is 8.8, which proves its high heterogeneity.

Obtained values (forest acreage) were verified by the AZP in some spots. For the study area, we analysed data concerning settlements from early middle ages to the late early modern period understood as the turn of the 18^th/19^th c. The archaeological data referred mainly to ceramic fragments found during field works. The material was used to determine the indicative range of fields cultivated in the area from the medieval to early modern period – and thus, approximate determination of the boundaries between arable land and the forest (Fokt 2012). In other words – the presence of ceramics around the settlements form the 16^th c. should indicate the presence of farmland before the period of so-called Dutch Colonization. In the case of ceramic fragments discovered around the Hollander settlements, this material did not play an important role in the process of forest reconstruction due to the potentially high risk of being mixed with fragments from other periods.

The above-mentioned analyses were supplemented by cartographic research method, i.e. the comparison of afforestation change between Gilly-Cron (1793) and Urmesstischblätter (approx. 1830). Scrubs from late 18^th c. were changed into clumps of trees in the beginning of 19^th c., which proves constant settlements’ development. Moreover, scrubs themselves can also be interpreted as the remains of the forest growing before the Hollander colonization. Almost full deforestation can be seen particularly in the centre of studied area (figure 6).

Fig. 5: Comparison of afforestation change between 1793 and 1830 on Gilly-Cron and Urmesstichblätter

The value of 10 hectares was adopted as the average area of the cleared forest by a single Hollander farm based primarily on the results achieved by geographical and geostatistical methods. The value combined with square-shaped buffers were the basis of the reconstruction map drawing. Forest type was reconstructed with the use of contemporary maps: topographic, soil, geological and potential natural vegetation of Poland, but also maps old maps, e.g. Urmesstischblätter shows information about the dominant vegetation type in the particular area. In this case, written sources were also considered (mostly the privileges granted to the settlers), as the type of forest existing before the location was mentioned there.

Literature

Fokt K. (2012). Późnośredniowieczne osadnictwo wiejskie na Dolnym Śląsku w świetle badań archeologicznych, Kraków.

Konias A. (2010), Kartografia topograficzna państwa i zaboru pruskiego od II połowy XVIII wieku do połowy XX wieku, Słups: Wydawnictwo Naukowe Akademii Pomorskiej.

Lorek D. (2011). Potencjał informacyjny map topograficznych Urmesstischblätter z lat 1822-33 z terenu Wielkopolski, Poznań: Zakład Graficzny UAM.

Matuszkiewicz W., Faliński J.B., Kostrowicki A.S., Matuszkiewicz J.M., Olaczek R., and Wojterski T. (1995). Potencjalna roślinność naturalna Polski. Mapa przeglądowa 1:300 000, Warszawa: Institute of Geography and Spatial Organization, Polish Academy of Sciences.

Medyńska-Gulij B. and Lorek D. (2008). Pruskie mapy topograficzne dla Wielkopolski do 1803 roku. Badania Fizjograficzne nad Polską Zachodnią. Seria A: Geografia Fizyczna 59:29–42.

Panecki T. (2014). Problemy kalibracji mapy szczegółowej Polski w skali 1:25 000 Wojskowego Instytutu Geograficznego w Warszawie. Polski Przegląd Kartograficzny 46(2):162–172.

Pelczyk A. (1996). Poolęderskie budownictwo mieszkalne na obszarze równiny nowotomyskiej. Studia Lednickie 4: 357-383.

Polish Territories of the Crown in the 16^th century. Spatial Database (2106). In digital form www.atlasfontium.pl

Scharfe W. (1972) Abriss der Kartographie Brandenburgs 1771–1821, Berlin, New York: Walter de Gruyter.

Słoń M. (ed.) (2014). Historical Atlas of Poland in the 2nd Half of the 16^th Century: Voivodeships of Cracow, Sandomierz, Lublin, Sieradz, Leczyca, Rawa, Plock and Mazovia ; Volume 2, Volume 3, Volume 4, Frankfurt am Main: Peter Lang.

State Archive in Poznań, Dokumenty Wiejskie 173, [English: Village documents 173]

State Archive in Poznań, Kościan Grodzkie 214, [English: Kościan Court Book 214]

State Archive in Poznań, Poznań Grodzkie 1204, 1205, 1207, [English: Poznań Court Book 1204, 1205, 1207]

Szady B. (2010). Granice powiatu łukowskiego w II połowie XVIII wieku - model liniowy i sieciowy, A. Barańska and W. Matwiejczyk (eds.), Narrata de fontibus hausta. Studia nad problematyką kościelną, polityczną i archiwistyczną ofiarowane Janowi Skarbkowi w siedemdziesiątą rocznicę urodzin, Lublin: RW KUL, 831–854.

Woyda S. (1975). Archeologiczne zdjęcie terenu – ogólne założenia metody w oparciu o doświadczenia mazowieckie, Wrocław.

---

Materiały do badań nad rekonstrukcją zalesienia na Równinie Nowotomyskiej w XVI w. autorzy, Instytut Historii im. Tadeusza Manteuffla Polskiej Akademii Nauk i Pracownia Geoinformacji Historycznej Katolickiego Uniwersytetu Lubelskiego jest dostępny na licencji Creative Commons Uznanie autorstwa - Bez utworów zależnych 4.0 Międzynarodowe.