AbstractArchival collections contain an underutilized wealth of biodiversity data, encapsulated in government files and other historical documents. In 1845, the Bavarian government conducted a comprehensive national survey on the occurrence of 44 selected vertebrate species across the country. The detailed expert responses from 119 forestry offices, totalling 520 handwritten pages, have been preserved in the Bavarian State Archives. In this study, we digitized, annotated, geographically referenced, and published these historical records, making them widely available as data for research and conservation planning. Our dataset, openly accessible through the Global Biodiversity Information Facility (GBIF) and Zenodo, contains 5,467 species occurrence records from 1845. Besides the binary presence/absence data, we have also published the original textual survey responses, which contain rich qualitative information, such as species abundances, population trends, habitats, forest management practices, and human-nature relationships. This information can be further processed and interpreted to address a range of questions in historical and contemporary ecology.
Background & Summary‘Die Wildkatze ist leider noch nicht gänzlich ausgerottet, und jedes Jahr liefert noch einige Exemplare.’ (Unfortunately, the wildcat has not yet been completely eradicated, and every year still yields a few specimens.) (Johann Mannert, head of the forestry office in Eltmann, Lower Franconia, 9 September 1845).Natural history museums1 and archival collections play a crucial role in preserving historical species records for biodiversity research and conservation, and documenting biodiversity changes throughout the Anthropocene. While the past two decades have seen rapid growth in digitizing museum specimens2, other historical records, especially textual sources of biodiversity remain largely untapped. These sources, typically found in “archives of societies”3, include administrative files, correspondence, maps, audiovisual records, and other forms of grey literature and media4. Unlike “archives of nature”, such as fossils, pollen, or soil core samples, and contemporary species occurrence records, “archives of societies” require specialized expertise in historical sciences to be discovered, meaningfully interpreted, and effectively used for ecological research. This challenge has led to the emergence of historical ecology—an interdisciplinary field bridging the biological sciences with societies’ archival records2,5.Some ecological studies have already utilized historical data such as those by Clavero et al.6 and by Nores and López-Bao7, both of whom used 19th-century geographical dictionaries to infer historical distributions of wolves in Spain, attempting to provide a baseline for current species conservation efforts, as did Clavero and Delibes for the lynx in Spain on a long time span8. Viana et al. harvested, analysed, and published data from a historical survey across 16th-century Spain9. Similarly, Govaerts drew on archaeological and historical sources of financial administration to document the occurrence of 46 bird species in late mediaeval Holland10. Another frequently used type of historical source is land and forest surveys, some of which date back to the 18th century. For example, Hanbery, Palik, and He compared an 1812 land survey with current forest surveys to examine shifts in forest community composition and biotic homogenisation in Minnesota, USA11. In Europe, Gschwantner et al. provide a review of historical forest inventories12. However, most literature on historical land and forest surveys focuses on tree communities and vegetation, with little reference to faunistic data. Furthermore, historical species records are rarely made accessible through biodiversity research infrastructures, such as the Global Biodiversity Information Facility (GBIF), except zoological and botanical collections. One possible reason for the lack of historical data is that, to be useful for biodiversity research, raw historical data must first be discovered, then understood, interpreted, annotated, and enriched in a process we refer to as “valorization”13.The dataset presented here was acquired from the Bavarian State Archives, resulting from interdisciplinary collaboration between natural sciences and humanities. We processed extensive historical sources on the occurrence status of 44 selected vertebrate species from a nation-wide standardized survey conducted in the Kingdom of Bavaria in 1845. The original source comprises the responses of 119 forestry offices, each contributing four handwritten pages. We transformed this traditional ecological knowledge—520 pages of handwritten text—into 5,467 classified, annotated, and geographically coded data points, now available on Zenodo14 and GBIF15 to support research in historical ecology (see Fig. 1 for an overview). In addition to binary species occurrence records (i.e. presence/absence), our dataset also includes transcribed qualitative responses from the forest officers who filled in the survey, as well as their names (see the opening quote for an example). These transcripts often provide valuable insights beyond simple presence or absence data.Fig. 1Forestry Offices and Rarest Species. The map displays the locations of all 119 forestry office seats as circles, with the circle diameters representing the number of species reported as present in each district. The ten rarest species in the survey are colour-coded by their presence. The map also includes the administrative districts of the Kingdom of Bavaria as they existed in 1845. The bottom right image shows a part of Wagner’s map from 1846. Map sources: own analysis; OpenStreetMap (OSM); QGIS; Bayerische Staatsbibliothek, BSB Cod.icon. 180 rc; HGIS Germany (https://hgl.harvard.edu/catalog/harvard-ghgis1848districts).Full size imageThe data are valuable for addressing a range of research questions in both contemporary and historical ecology, meeting the critical need for historical species occurrence records and long-term time series that help better understand biodiversity changes in the Anthropocene16. Historical species records, when critically assessed, can inform conservation baselines, national red lists, and policy-making17. The quote at the beginning of this article illustrates how qualitative data also offer insights into historical human-nature relationships. Additionally, these records provide information on changing species habitats, precise locations of animal sightings, and possible reasons for population declines and local extinctions. Being informed from the past18 is “vitally important”19 for various ecological applications, including restoration and conservation efforts, where historical baselines are essential8.Aside from a report by the conductor of the survey and a brief mention in a biography20, the survey and its data have received little attention in the literature. Neither a systematic evaluation of the survey nor a publication of the original sources has been undertaken to date.MethodsStudy areaThe data presented here encompass the entire Kingdom of Bavaria, then an independent nation; now, with some territorial changes, the largest federal state in Germany. This area includes a substantial territory (76,770 km2, of which 24,560 km2, or 31.99%, are woodland), almost twice the size of the Netherlands, and a variety of ecosystems, representing 21 of the 24 landscape types currently described by the German Federal Agency for Nature Conservation (https://www.bfn.de/daten-und-fakten/biogeografische-regionen-und-naturraeumliche-haupteinheiten-deutschlands). These include smaller parts of the Austrian Limestone Alps as well as left-bank Rhine regions, such as the Palatinate Forest adjacent to the Vosges Mountains along the French-German border. In 1845, the area was predominantly agricultural and pre-industrial with an overall population density of 57.8 inhabitants per km2 21.Data processing (Overview)Data processing (see Fig. 2) encompassed the following steps, which are described in more detail below:
Source discovery and historical contextualization
Digitization and creation of machine-readable texts
Datafication
Transformation and publication
Fig. 2Data Processing Workflow (Overview). This diagram illustrates the workflow used to create the Historical Animal Observation Records by Bavarian Forestry Offices (1845) dataset, from the discovery of the archival sources to the publication of the final data in public repositories. The top row describes each processing step, while the bottom row outlines the resulting outputs.Full size imageFigure 3 shows an entry of a sample record in its original form from the Tegernsee (Salforste) office, describing the occurrence of the Eurasian lynx (Lynx lynx). The text is presented in both the original language and its English translation:Fig. 3Sample Response. This image is part of the Tegernsee (Salforste) report (cf. Figure 4 for the full-page image), documenting the presence of the lynx. The transcribed original language, along with its English translation, is provided in the text. Source: BayHStA Zool. Staatssammlung, 217, p. 14.Full size imageLuchs: “Seitdem /: im Jahre 1826:/ ein ganzes aus den alten u. zwey jungen Luchsinnen bestehendes Gehecke am Hirschberge ausgerottet worden, sind nur noch Einzelne dieser Raubthiere diese in den hiesigen Gebirgen erschienen, früher zahlreich, jedoch immer nur periodisch”Lynx: “Since /: in 1826:/ an entire enclosure consisting of the old and two young female lynxes on the Hirschberg was eradicated, only individuals of these predators have appeared in the local mountains, formerly numerous, but always only periodically”Source discovery and historical contextualizationThis study is based on historical documents from 1845 containing animal observations recorded by Bavarian forestry offices. Foresters used predefined, standardized forms sent to them to document the presence or absence of 44 listed animal species within their districts, as well as fish species in local waters. In some cases, foresters provided additional information on species beyond those listed on the forms. This source is particularly valuable due to its comprehensive coverage of all of Bavaria at a specific point in time—a rarity among archival sources. Observations were documented by a total of 119 forestry offices across the eight Bavarian government districts. The documents are organized into ten files, with one file for each of the eight government districts (Upper Bavaria, Lower Bavaria, Upper Franconia, Middle Franconia, Lower Franconia, Upper Palatinate, Palatinate, Swabia-Neuburg), a file for the Salforste region—a forested area located in the border region between Germany and Austria, historically managed as a shared resource between Bavaria and Austria—and a summary for all of Bavaria. Together, these documents comprise approximately 520 pages.The documents are housed in the Bavarian State Archives as part of the record group for the Zoological State Collection (https://www.gda.bayern.de/service/findmitteldatenbank/Kapitel/0ea38d12-d425-4b3e-a497-7b6830f439e1). This record group, spanning the period from 1827 to 1990, consists of 1,442 archival units. These units include documents related to the organisation, administration, and personnel matters of the Zoological State Collection, along with records of acquisitions, inventory catalogues, professional staff correspondence, and scientific research. The animal occurrence records completed by the forestry offices were originally held in the Zoological State Collection in Munich and were transferred to the Bavarian State Archives in 2013.The context of these documents is deeply intertwined with the complex history of the Zoological State Collection. Established in 1827 as an independent public research institution by order of King Ludwig I, the Zoological Collection was initially separated from the scientific collections of the Bavarian Academy of Sciences. However, it soon became dependent on Ludwig Maximilian University, which had to transfer its zoological collection to the state cabinet. The collection’s director, or conservator, was typically the university’s Professor of Zoology. Simultaneously, the “General Conservatory of State Scientific Collections” was founded, intended to secure the collections’ independence. However, this agency’s effectiveness was limited by the Academy’s influence, as its president also served as the General Conservator. As a result, the zoological cabinet existed in an ambiguous organizational and personnel structure, with its scientific staff primarily sourced from the university, but required approval from the state collection administration, which was heavily reliant on the Academy’s resources.The valuable dataset presented here originated from a commission by Bavarian Crown Prince Maximilian II, who tasked the zoologist Johann Andreas Wagner with mapping the distribution of Bavaria’s most significant animal species22. Wagner, who had been Professor of Zoology at the University of Munich and Deputy Curator of the State Zoological Collection since 1836, worked diligently to expand the collection and made notable contributions to zoological taxonomy20. To fulfil the prestigious royal commission and ensure a comprehensive survey of Bavaria’s animal population, Wagner decided to involve the royal forestry offices22. These offices were likely selected because the foresters were not only expected to possess a thorough knowledge of local wildlife but were also regarded as qualified experts due to their academic training.In 1845, the 119 royal forestry offices, then subordinate to the Bavarian Ministry of Finance, were issued a directive from the Ministry, mandating that they systematically record the animal species present in their respective forestry districts. The motivation behind the survey is evident from the official letter: foresters were explicitly asked to document the animals from a “scientific perspective” (cf. Figure 4). This initiative arose during a period of rapid scientific and economic development. Concurrently, there was a growing awareness of the relationship between humans and nature, with particular interest in the systematic and descriptive documentation of the natural world. While the geographical distribution of plants had already been well studied, Wagner observed that zoology still lagged behind in this regard23.Fig. 4Royal Order and Tegernsee Response. Left: Royal order from 13 August 1845: “Zur Kenntniß der geographischen Verbreitung der Thier- und Baumarten in Bayern — vom wissenschaftlichen Standpunkt aus gesehen — wird gewünscht, daß die in hinreichender Anzahl beigefügten Schemata, Litera A. und B., von den K. Forstämtern ausgefüllt werden.” (For an understanding of the geographical distribution of animal and tree species in Bavaria—from a scientific perspective—it is desired that the attached forms, Litera A and B, be completed by the Royal Forestry Offices.) Source: BayHStA, Zoologische Staatssammlung, 208. Right: Completed first page of the survey form from the Salinenforstamt Tegernsee. Source: BayHStA Zoologische Staatssammlung, 217.Full size imageThe survey responses, comprising more than 5,400 entries in short prose, offer insights not only into species occurrence but also into habitats, their changes, and the effects of human activity on wildlife populations. For example, the survey contains the last recorded sightings of the beaver (Castor fiber) in Bavaria before its extinction around 186724. Wagner anticipated the beaver’s impending extinction, attributing it to particularly lucrative but officially prohibited hunting22.The survey results formed the basis for a map visualizing animal distribution in Bavaria, as commissioned by the Crown Prince25. Andreas Wagner also presented these findings in a lecture at the Bavarian Academy of Sciences on 21 February 1846, where he had been a member since 1835. This lecture was subsequently published in Gelehrte Anzeigen, a journal issued by members of the Bavarian Academy of Sciences22.Digitization and creation of machine-readable textsIt is typical for many historical sources found in archives of societies that the raw data of the 1845 survey consist of paper-based archival files in which handwritten texts represent the information of interest. To enable further processing, we created digital images of these original documents following the German Research Council (DFG) recommendations26, specifically uncompressed TIFF images with a resolution of 300 ppi (pixels per inch) as master files to allow future reuse and upcycling27. We then transformed these image-based raw data into machine-processable texts through a two-step process: first, layout recognition (identifying regions and script baselines), and, second, content transcription using AI-based Handwritten Text Recognition (HTR) via the pre-trained Transkribus model “The Text Titan”, without additional custom training28. This automated process produced an initial Character Error Rate (CER) of approximately 7%, which was significantly improved through manual post-correction by trained palaeographers. The texts were transcribed “as is”—preserving the original language, maintaining spelling errors, and leaving abbreviations unresolved—to keep the transcriptions as faithful to the originals as possible.DataficationThese raw texts were organized by forestry office name and taxon name, creating a total of 5,467 entries for 44 predefined species, along with additional species reported by foresters. We interpret this processing of a non-digital artefact into computable codes as datafication27, whereby data of interest were extracted from textual content that was originally unstructured and not formatted for direct use. Instead, the information followed a free-text format, adhering to the instructions provided in Wagner’s questionnaire (presented here in English translation):“List of animal species of whose existence and place of residence information is desired. Note: If the distribution is limited to certain localities only, the nearest village or locality should be included in parenthesis. In the case of birds, only those which breed in the district itself or which spend the winter there, or which are currently the subject of hunting, should be listed. – It should also be stated whether the species is common or rare.”We applied the following rules for datafication:
1.
Metadata (per district): For each of the 119 forestry districts, the location of its office seat was researched and geographically coded. Example: “Forstverwaltung Deggendorf” → office seat Deggendorf, Lower Bavaria → longitude = 12.9603, latitude = 48.8348. Sources: survey headers, State Directory Kingdom of Bavaria 184529, and OpenStreetMap (OSM).
2.
Species names: Mapping of the predefined 44 species names and an additional 91 individually reported, partly historical, names to their scientific names and taxonomy (see Tables 1, 2). The trivial names and Latin names used in the historical dataset serve as the central point of integration for the data into the service infrastructures of biodiversity informatics, e.g., for the analysis of historical species distributions or changes in biodiversity patterns over time. However, for the interoperability of the data, it is necessary to understand the species concepts used and to link them to recent concepts using stable persistent identifiers. Beyond clarifying the concepts, this also avoids the problem of orthographic variants of species names during data integration. We used the checklist infrastructure of the German Federal Agency for Nature Conservation (BfN) for the taxonomic annotation of the data (https://checklisten.rotelistezentrum.de/api/public/swagger-ui). On the one hand, it provides the taxonomic and geographical coverage necessary for the annotation of forestry data. On a second level, the infrastructure allows taxonomic concepts to be precisely referenced using persistent identifiers, providing the basis for the semantic annotation of the taxon names used30. Example: “Halbente” → Knäkente → anas querquedula. Sources: survey, checklist (e.g., https://checklisten.rotelistezentrum.de/api/public/1/taxon/24804); historical literature and dictionary in rare cases (e.g., https://www.dwds.de/wb/dwb/halbente).Table 1 Species Overview (mammals).Full size tableTable 2 Species Overview (birds and reptiles).Full size table
3.
Text: Text is transcribed character by character as is in the original source including abbreviations and potential misspelling. Unsure transcriptions are marked as ‘#…#’. A significant amount of entries repeat the previous line by using ‘ditto’ words or signs, or they aggregate entries by using curly brackets. Occurrence data have been drawn from the line referred to by the signs or brackets in these cases. Example: “Lynx: Kömmt nicht vor; Wild cat: dito; Wolf: dito” yields occurrence = 0 for all three species (Deggendorf, Lower Bavaria).
4.
Occurrences: Binary classification (absence or presence) of occurrences per species and district. This basic classification of occurrence also caters for the fact that the responses by the forestry offices were quite diverse in scope and quality. On the lower end of the scale of detail, an entry could simply read “[species] does not exist here” or even only a strike-through of the species names (both classified as ‘0’) while on the upper end, detailed information about habitat or specific location could have been given. In cases of reference to additional species or more comprehensive reporting such as specific time of observation, a new entry has been created. Source: survey.
a.
Classification as 1: We could draw from the texts that the reporting office was sure or at least provided good reason that a specimen of this species had recently (in 1845) been seen within the boundaries of its district. We did not distinguish the type of the animal’s visit (e.g. resident or migratory bird); nor have statements about quantity and abundance (yet) been datafied (e.g. rare occurrence, close to extinction, or regular). Where such information is available, future research may extract it from the texts given in the dataset. Example: “The hamster is found only very rarely.” (Freysing, Lower Bavaria).
b.
Classification as 0: There is no or no justifiable evidence for an observation. Example: “The bear is not present.” (Freysing, Lower Bavaria). This also covers the few cases in which an occurrence is reported to be situated outside a natural environment (in game parks or pheasantries). Example: “The pheasant can only be found in pheasantries” (Bayreuth, Oberfranken).
5.
Time: We generally assume all reports refer to the year 1845 or only few years earlier if not otherwise stated. Where otherwise stated, the given date has been coded and used as time of observation. In case of the latter, a new data entry has been created (see above). Examples: “has not occurred since 1710, when the last one was seen in the former Rehau district and caught in the Sparneck district” — > creates two entries, one for 1710 (classified ‘1’), one for 1845 (classified ‘0’). Source: survey.
6.
Location: We generally assume all occurrences are located within a default radius of 20 km from the seat of the forestry office.
Transformation and publicationFinally, all data were transformed using a standardized terminology based on the Access to Biological Collection Data (ABCD) Schema (http://www.tdwg.org/standards/115). This transformation can also be seen as “valorisation”13—creating value and impact from the historical sources, the knowledge they contain, and the historian’s expertise required for their mobilisation. We have published the data under a CC BY licence in two formats
1.
A dataset uploaded to the Global Biodiversity Information Facility (GBIF) and published via the GBIF data services and the GBIF-hosted portal Lebendiger Atlas der Natur Deutschlands (LAND, https://land.gbif.de). Here, the data can be used within the given framework and its retrieval and visualization tools, and in the broader contexts of the entire GBIF database, allowing for spatial and temporal analysis as well as species comparisons15.
2.
A dataset uploaded to the online repository Zenodo, available for download for in-depth analysis using statistical and text mining tools and Geographic Information Systems (GIS). This format also supports cross-disciplinary applications14.
Data RecordsThe following data records are available:
1.
Digital images from the original archival sources (https://www.gda.bayern.de/service/findmitteldatenbank/Kapitel/0ea38d12-d425-4b3e-a497-7b6830f439e1);
2.
A dataset available at Zenodo, containing14:
Transcription data (as XML file);
Main occurrence data table (as CSV file);
Species names data table (as CSV file);
Forestry offices and administrative districts data tables (as CSV file);
Additional information data table (as CSV file).
High-resolution digital images of all original source files from the digitization process are accessible via the online service of the Bavarian State Archives, organized in Bavarian administrative districts and including the file for the Salforste region. The images are displayed in the original order of the files, with full navigation and image manipulation options available through the DFG Viewer (https://dfg-viewer.de/). Users can also utilise the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) interface for downloading or linking metadata. To enable verification of the transcriptions, the main data table includes a direct URL linking to the specific digital image corresponding to each entry.The data repository includes the transcription results in both human-readable and machine-readable XML formats. Metadata in this file align with the archival metadata used by the Bavarian State Archives in their catalogues and the online archive. Metadata fields follow the standard of the Text Encoding Initiative (TEI, https://tei-c.org). The file content maintains the original order and is hierarchically organized as follows: administrative district → forestry office → species record. Text fields present entries “as is,” with no interpretation, correction, annotation, or analysis. These transcriptions allow users to verify each entry in the main data table without requiring palaeographic expertise.The main occurrences data table contains:
5,467 records of observation data in structured CSV format, following the hierarchical structure of administrative district → forestry office → species with one row per species and forestry office;
Additional rows for special cases: reported extra species, earlier (than 1845) observations reported;
Transcriptions of the textual evidence for each entry (see Table 3 for a content breakdown);Table 3 Qualitative breakdown of textual entries.Full size table
Direct links to the digital images carrying the textual evidence;
Identifiers linking each entry to the respective species and forestry office tables.
The species names data table contains:
Species names in historical writing, English translation, and modern scientific names;
Direct references to the German Red List checklist dataset.
The forestry offices and administrative districts data tables contain:
Names and location (longitude and latitude) of the seats of each of the 119 forestry offices to allow approximate geographical coding of each observation;
Names and English transcriptions of the eight administrative districts of the Kingdom of Bavaria in 1845.
Additional information about the forestry offices and officers, sourced from the 1845 Bavarian State Directory, is provided in a separate table29. For peer review, version 1.3 of the dataset has been used (https://zenodo.org/records/14008158).Technical ValidationThe data presented here are derived from text-based historical sources, which means that the resulting dataset represent the original observations indirectly in two respects31: first, we did not observe the animals ourselves; and second, we did not directly question the foresters who reported these observations. Instead, as is typical and necessary for historical data drawn from societies’ archives, we critically relied on the traces32 left behind by foresters in their written reports from 1845. This approach raises methodological questions and challenges, closely tied to questions of data validation33. As with any work in historical ecology, an interdisciplinary approach was necessary and was applied to ensure reliable and verifiable data.To validate the historical information found in the original documents, we employed the widely accepted methods of historical source criticism. Overall, there is no doubt that these sources are authentic and accurately represent what they claim to be.We vetted the following:
1.
Origin of responses: The survey documents were exchanged via the official postal service, some responses were hand-signed. The names of the signatories match those in the official register of forest managers in the 1845 State Directory of the Kingdom of Bavaria29.
2.
Integrity and completeness of the archival record: The authorities have well-documented the provenance of these files, and the sources show no indication of later manipulations. A response exists for each forestry office listed in the State Directory29, suggesting that no entries are missing.
3.
Dating: There is no doubt that the survey took place in 1845. The ministry’s executive order dates to that year, the names of forestry officers in the responses align with those in the State Directory issued annually29, and Wagner’s 1846 report to the Academy confirms the survey’s timing.
4.
Trustworthiness of content – intentional errors: While individuals might have had motives to misrepresent or conceal information (e.g., involvement in wildlife poaching), it is highly unlikely for forestry officers, who were royal officials, to act deceptively. Given their duty to uphold official responsibilities without personal gain, intentional misreporting is improbable.
5.
Trustworthiness of content – unintentional errors: Forestry officers possessed sound knowledge of fauna and flora, as they were required by law to undergo three years of specific training, including zoology, at a renowned university-level institution in Aschaffenburg or Munich34. We verified this for individual officials using their personnel files in the State Archives.
6.
Variability in response quality: As noted, there are significant differences in the quality and scope of the responses, suggesting that foresters provided answers with varying levels of commitment. This variability imposes limitations on the comparability of responses, particularly in cases involving sightings of rare animal species.
Secondly, we validated data processing through the following methods:
1.
Transcriptions were double-checked, and in cases of uncertainty, triple-checked by independent reviewers. Names of individuals, forest districts, and sub-districts were verified against the 1845 State Directory29, while place names were cross-referenced using historical gazetteers, regional websites, and historical maps. Unclear readings were marked accordingly. All transcriptions can be validated against the digitized originals, with references provided in the dataset.
2.
Binary occurrence classifications were independently double-checked. These classifications can be verified at any time using the texts within the dataset.
3.
Species names and classifications were verified, with rare, uncertain cases discussed and resolved based on the likelihood of species occurrence in Central Europe.
4.
Consistency and completeness checks were automatically performed by script for each dataset version. This script validated the internal logic of entities and relationships by:
a.
Ensuring that each entry in AOD_Observation_Record_Data.csv has matching values for OfficeName and District in AOD_Offices.csv;
b.
Ensuring that each SpeciesName in AOD_Observation_Record_Data.csv corresponds to a valid entry in AOD_Species_Names.csv;
c.
Checking that each Date in AOD_Observation_Record_Data.csv falls within a reasonable range (1600–1845);
d.
Validating AOD_transcriptions.xml against its schema and ensuring that each core observation entry in AOD_Observation_Record_Data.csv has a corresponding XML element in AOD_transcriptions.xml.
Usage NotesThe data presented here can be regarded as scientifically influenced traditional ecological knowledge35, with studies using this data classified as applied historical ecology—that is, “the use of traditional knowledge (TK) to help address problems of people and the environment today”36. As archives of societies likely hold extensive additional knowledge, such as records from past foresters—our approach to mobilising these sources may serve as a model for similar projects worldwide.When using this dataset for research, it is essential to consider typical characteristics of historical data. Here are specific points to keep in mind:Geographical information
Changing Boundaries and Land Use: Just as land use, habitats, and ecosystems have changed over time, so too have administrative boundaries. Therefore, geographic analysis requires using maps and shapefiles that accurately reflect the relevant period, such as those provided by GeoPortal Bayern (https://geoportal.bayern.de/geoportalbayern/suche/suche?0&q=historisch) or HGIS Germany (https://hgl.harvard.edu/catalog/harvard-ghgis1848districts).
Geographic Resolution of Observations: The geographic resolution in this dataset is based on the administrative seats of forestry offices, giving each observation an average coverage area of 645 km2. The Salforste region is a special case: due to a bilateral treaty between Bavaria and Austria (the Salinenkonvention of 1829 which remains in force today), the jurisdiction—and thus the observations—of Bavarian forestry offices extend into parts of Austria37.
Detailed Location Information: Some observations contain more specific location details. To make use of this higher geographic resolution at the data-entry level, geographic coding of the qualitative textual information is required, which is best supported by historical gazetteers and maps Local knowledge may also be helpful in certain cases.
Time considerations
Observation Timing: “1845” does not necessarily indicate a precise observation date; rather, it reflects a forester’s assessment in summer 1845 that the species in question was generally present (or absent) within his jurisdiction.
Species and taxonomy
Classification Limitations: While we have classified the observation records to the best of our knowledge, there may occasionally have been misunderstandings by foresters about the species being surveyed. In such cases, interpreting the original text may be necessary, especially since foresters often explained their interpretations of the questionnaire.
Additional Species Mentions: Mentions of additional species by foresters are uneven and unsystematic. Unlike the primary 44 species, an absence of reporting on these species does not imply their absence; rather, it simply means no information was provided.
Ducks Family Reporting: Ducks were surveyed as a family, resulting in varied responses, from general statements (“ducks occur”) to detailed species lists. For general mentions, absence of notes on specific species (e.g., Anas platyrhynchos, mallard) does not imply the species did not occur, but rather reflects a lack of specific information.
Special Cases – SP_0023 (“Alpendohle”) and SP_0024 (“Steinkrähe /: corvus graculus:/”): Differentiation between these two terms may be inconsistent in the data, as some foresters appeared uncertain. “Steinkrähe,” clarified by Wagner as “corvus graculus,” is no longer common in German (cf. Figure 5) and may historically refer to the red-billed chough (Pyrrhocorax pyrrhocorax)38. To approximate accurately, consider the red-billed chough (“Alpenkrähe” in modern German, https://www.gbif.org/species/2482552) and yellow-billed chough (“Alpendohle”, https://www.gbif.org/species/2482553), though differentiation within the responses may not be consistent.Fig. 5Usage of Species Names (1800–2022). Historical usage trends for the terms “Steinkrähe” (blue), “Alpenkrähe” (red), and “Alpendohle” (green) show that all three terms were used relatively evenly (though not synonymously) around 1845. Since then, “Steinkrähe” has become uncommon, leaving only “Alpendohle” (yellow-billed) and “Alpenkrähe” (red-billed) in common use. Source: Google Books Ngram Viewer (https://books.google.com/ngrams/).Full size image
Qualitative information
Historical Language Considerations: Keep in mind that language, and word meanings in particular, change over time. For example, the German term “merkwürdig” meant “notable” in the 19th century, but now means “bizarre.”
Subjectivity and Context: The textual data provide subjective statements that need historical contextualisation and interpretation, particularly regarding baseline conditions39. Terms like “häufig” (frequent, often) may have different implications among respondents and may have had shifted in meaning and their baselines over time40.
The dataset opens up research questions beyond statistical analysis, including:
Influences on Species and Ecosystems: Foresters reported not only species occurrences but also—though inconsistently—on factors affecting species’ living conditions, such as human actions, landscape changes, and climate influences41. Since the survey took place at the end of the European Little Ice Age, this contextual information could be particularly relevant. The dataset also provides insights into human attitudes towards animals—such as hunting—along with detailed habitat and location data. This content offers potential for studies in geographic anthropology and historical biology. To support such research, further semantic enrichment and annotation of the data would be beneficial.
Comparative and Combined Analyses: The dataset can be compared or integrated with similar datasets from other regions or periods. Additionally, further research is needed to develop automated methods for data mobilisation and analysis of archival records42.
Advancing Biodiversity Informatics: This work contributes to mobilising historical data that can be used to analyze ecosystem patterns and species distributions, as well as to explore underlying causes. Semantic annotation of species names with stable, persistent identifiers is an essential step towards integrating this dataset into a global biodiversity knowledge graph43. To enhance integration within the interconnected field of biodiversity informatics, further annotation of data on individuals, habitat types or other critical concepts could provide a foundation for improved automated data integration and inference44,45.
Code availability
No custom code has been used.
ReferencesSanders, N. J., Cooper, N., Davis Rabosky, A. R. & Gibson, D. J. Leveraging natural history collections to understand the impacts of global change. J Anim Ecol 92, 232–236 (2023).Article
PubMed
Google Scholar
Nelson, G. & Ellis, S. The history and impact of digitization and digital data mobilization on biodiversity research. Phil. Trans. R. Soc. B 374, 20170391 (2019).Article
Google Scholar
Brönnimann, S., Pfister, C. & White, S. Archives of Nature and Archives of Societies. in The Palgrave handbook of climate history (eds. Mauelshagen, F., Pfister, C. & White, S.) 27–36 (Palgrave Macmillan, London, 2018).Decocq, G. A General Introduction to Historical Ecology. in Historical Ecology: Learning from the Past to Understand the Present and Forecast the Future of Ecosystems (ed. Decocq, G.) 1–9 (Wiley, 2022).Vellend, M., Brown, C. D., Kharouba, H. M., McCune, J. L. & Myers-Smith, I. H. Historical ecology: using unconventional data sources to test for effects of global environmental change. Am J Bot 100, 1294–305 (2013).Article
PubMed
Google Scholar
Clavero, M., García-Reyes, A., Fernández-Gil, A., Revilla, E. & Fernández, N. On the misuse of historical data to set conservation baselines: Wolves in Spain as an example. Biological Conservation 276, 109810 (2022).Article
Google Scholar
Nores, C. & López-Bao, J. V. Historical data to inform the legal status of species in Europe: An example with wolves. Biological Conservation 272, 109639 (2022).Article
MATH
Google Scholar
Clavero, M. & Delibes, M. Using historical accounts to set conservation baselines: the case of Lynx species in Spain. Biodivers Conserv 22, 1691–1702 (2013).Article
MATH
Google Scholar
Viana, D. S., Blanco-Garrido, F., Delibes, M. & Clavero, M. A 16th-century biodiversity and crop inventory. Ecology 103, 3783 (2022).Article
MATH
Google Scholar
Govaerts, S. Biodiversity in the Late Middle Ages: Wild Birds in the Fourteenth-Century County of Holland. Environment and History 30, 241–266 (2024).MATH
Google Scholar
Hanberry, B. B., Palik, B. J. & He, H. S. Comparison of historical and current forest surveys for detection of homogenization and mesophication of Minnesota forests. Landscape Ecol 27, 1495–1512 (2012).Article
Google Scholar
Gschwantner, T. et al. Growing stock monitoring by European National Forest Inventories: Historical origins, current methods and harmonisation. Forest Ecology and Management 505, 119868 (2022).Article
Google Scholar
Perez Vico, E., Sörlin, S., Hanell, L. & Salö, L. Valorizing the Humanities. in Making Universities Matter: Collaboration, Engagement, Impact (eds. Mattsson, P., Perez Vico, E. & Salö, L.) 211–232 (Springer Nature Switzerland; Imprint Springer, Cham, 2024).Rehbein, M. et al. Historical Animal Observation Records by Bavarian Forestry Offices (1845). https://doi.org/10.5281/zenodo.13899541 (2024).Rehbein, M. et al. Historical Vertebrate Distributions in Bavaria 1845. Bayerisches Hauptstaatsarchiv https://doi.org/10.15468/UKHF9P (2024).Dornelas, M. et al. BioTIME: A database of biodiversity time series for the Anthropocene. Glob Ecol Biogeogr 27, 760–786 (2018).Article
PubMed
PubMed Central
MATH
Google Scholar
Braje, T. J. & Rick, T. C. From forest fires to fisheries management: anthropology, conservation biology, and historical ecology. Evol Anthropol 22, 303–11 (2013).Article
PubMed
Google Scholar
Safford, H. D., Hayward, G. D., Heller, N. E. & Wiens, J. A. Historical Ecology, Climate Change, and Resource Management: Can the Past Still Inform the Future? in Historical Environmental Variation in Conservation and Natural Resource Management (eds. Wiens, J. A., Hayward, G. D., Safford, H. D. & Giffen, C. M.) 46–62. https://doi.org/10.1002/9781118329726.ch4 (Wiley, 2012).Higgs, E. et al. The changing role of history in restoration ecology. Frontiers in Ecol & Environ 12, 499–506 (2014).Article
MATH
Google Scholar
Alaoui Soulimani, A. Naturkunde Unter Dem Einfluss Christlicher Religion: Johann Andreas Wagner (1797 - 1861): Ein Leben Für Die Naturkunde in Einer Zeit Der Wandlungen in Methode, Theorie Und Weltanschauung. (Shaker, Aachen, 2001).Die Forstverwaltung Bayerns. (Hennings und Hopf, Erfurt, 1845).Wagner, A. Beyträge zur Kenntniß der bayerischen Fauna. Gelehrte Anzeigen, herausgegeben von Mitgliedern der königlich bayerischen Akademie der Wissenschaften 22, 649-656657-664665-672673-680697–700 (1846).Wagner, A. Die geographische Verbreitung der Säugethiere. Abhandlungen der mathematisch-physikalischen Klasse der königlich bayerischen Akademie der Wissenschaften 1, 1–146 (1844).MATH
Google Scholar
Zahner, V. Der Biber in Bayern: Eine Studie Aus Forstlicher Sicht. (1997).Wagner, A., Weber, J. C. & v. Massenbach, F. Uebersichtskarte der Verbreitungsverhältnisse der merkwürdigsten wildlebenden Thiere in Bayern: Ein Versuch nach Angabe Königlicher Forstämter und eigener Erfahrung. (1846).Altenhöner, R. et al. DFG-Praxisregeln ‘Digitalisierung’. Aktualisierte Fassung 2022. https://zenodo.org/records/7435724 (2022).Donig, S. & Rehbein, M. Für eine „gemeinsame digitale Zukunft“. Eine kritische Verortung der Digital History. Geschichte in Wissenschaft und Unterricht 72, 527–545 (2022).MATH
Google Scholar
Kahle, P., Colutto, S., Hackl, G. & Mühlberger, G. Transkribus - A Service Platform for Transcription, Recognition and Retrieval of Historical Documents. in 2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) 19–24. https://doi.org/10.1109/ICDAR.2017.307 (IEEE, 2017).Hof- und Staatshandbuch des Königreichs Bayern (1845).Hüllbusch, E. et al. A Concept-Aware API for the Taxonomic Checklist Infrastructure in Germany. Biodiversity Information Science and Standards 8, (2024).Armstrong, C. G. & Junqueira, A. B. Conclusion: Historical ecology: challenges and perspectives in a changing world. in Methods in historical ecology: Insights from Amazonia (eds. Odonne, G. & Molino, J.-F.) 171–177 (Routledge, Abingdon, Oxon; New York, NY, 2021).Wylie, A. Rock, Bone, and Ruin: A Trace-centric Appreciation. Philosophy, Theory, and Practice in Biology 11, (2019).Edmonds, M. The Pleasures and Pitfalls of Written Records. in The historical ecology handbook (eds. Egan, D. & Howell, E. A.) 73–99 (Island Press, Washington, DC, 2001).Hamberger, J. & Bauer, O. Wald, Mensch, Heimat: Eine Forstgeschichte Bayerns. (Laubsänger-Verlag, Freising, 2019).McGregor, D. Linking Traditional Ecological Knowledge and Western Science: Aboriginal Perspectives from the 2000 State of the Lakes Ecosystem Conference. The Canadian Journal of Native Studies 28, 139–158 (2008).MATH
Google Scholar
Balée, W. & Dudley, M. Applied Historical Ecology. in Methods in historical ecology: Insights from Amazonia (eds. Odonne, G. & Molino, J.-F.) 162–170 (Routledge, Abingdon, Oxon; New York, NY, 2021).Wegmaier, A. Salinenkonvention 1829 und 1957. in Historisches Lexikon Bayerns (2006).Schlegel, H. Kritische Übersicht Der Europäischen Vögel: = Revue Critique Des Oiseaux d’Europe. (Arnz, Leiden, 1844).Clavero, M., García‐Reyes, A., Fernández‐Gil, A., Revilla, E. & Fernández, N. Where wolves were: setting historical baselines for wolf recovery in Spain. Animal Conservation 26, 239–249 (2023).Article
Google Scholar
Haider, T., Perschl, T. & Rehbein, M. Quantification of Biodiversity from Historical Survey Text with LLM-based Best-Worst Scaling. Preprint at https://doi.org/10.48550/arXiv.2502.04022 (2025).Rehbein, M. From Historical Archives to Algorithms: Reconstructing Biodiversity Patterns in 19th Century Bavaria. Preprint at https://doi.org/10.20944/preprints202503.1156.v1 (2025).Rehbein, M. On Computational Historical Ecology, Part I. Computational Historical Ecology https://che.hypotheses.org/69 (2024).Page, R. D. M. Ozymandias: a biodiversity knowledge graph. PeerJ 7, 6739 (2019).Article
MATH
Google Scholar
Groom, Q. et al. People are essential to linking biodiversity data. Database (Oxford) 2020, (2020).Güntsch, A. et al. A botanical demonstration of the potential of linking data using unique identifiers for people. PLOS ONE 16, 0261130 (2021).Article
MATH
Google Scholar
Download referencesAcknowledgementsWe are grateful to the General Directorate of the Bavarian State Archives for providing access to the historical sources. We also thank the community of the German National Research Infrastructure on Biodiversity for valuable discussions, Katja Luther and Jörg Holetschek for assistance with the GBIF publication, and Robert Forkel for support in validating the data. The authors acknowledge support by the Open Access Publication Fund of University Library Passau. The project is part of the German National Research Infrastructure on Biodiversity (NFDI4Biodiversity) which is supported by the German Research Foundation DFG under the grant agreement number 442032008.FundingOpen Access funding enabled and organized by Projekt DEAL.Author informationAuthors and AffiliationsChair of Computational Humanities, University of Passau, Passau, GermanyMalte Rehbein, Bettina Haas & Tobias PerschlFreie Universität Berlin, Botanic Garden and Botanical Museum Berlin. Center for Biodiversity Informatics and Collection Data Integration (ZBS), Berlin, GermanyBelen Escobari & Anton GüntschResearch Institute for Farm Animal Biology (FBN) Dummerstorf, Dummerstorf, GermanySarah FischerGeneraldirektion der Staatlichen Archive Bayerns, Munich, GermanyGiada Matheisen & Alois WieshuberInstitute of Biodiversity, Friedrich Schiller University Jena, Jena, GermanyThore EngelDepartment of Biodiversity and People, Helmholtz Centre for Environmental Research - UFZ, Leipzig, GermanyThore EngelGerman Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, Leipzig, GermanyThore EngelAuthorsMalte RehbeinView author publicationsYou can also search for this author inPubMed Google ScholarBelen EscobariView author publicationsYou can also search for this author inPubMed Google ScholarSarah FischerView author publicationsYou can also search for this author inPubMed Google ScholarAnton GüntschView author publicationsYou can also search for this author inPubMed Google ScholarBettina HaasView author publicationsYou can also search for this author inPubMed Google ScholarGiada MatheisenView author publicationsYou can also search for this author inPubMed Google ScholarTobias PerschlView author publicationsYou can also search for this author inPubMed Google ScholarAlois WieshuberView author publicationsYou can also search for this author inPubMed Google ScholarThore EngelView author publicationsYou can also search for this author inPubMed Google ScholarContributionsMalte Rehbein: lead scientist, conceptualisation, statistical analysis, GIS-related work, main author; Belen Escobari: taxonomy; Sarah Fischer: review; Anton Güntsch: taxonomy; Bettina Haas: transcriptions, archival research; Giada Matheisen: co-author, archival research; Tobias Perschl: transcriptions; Alois Wieshuber: archival research; Thore Engel: co-author, conceptualisation.Corresponding authorCorrespondence to
Malte Rehbein.Ethics declarations
Competing interests
The authors declare no competing interests.
Additional informationPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Reprints and permissionsAbout this articleCite this articleRehbein, M., Escobari, B., Fischer, S. et al. Quantitative and qualitative Data on historical Vertebrate Distributions in Bavaria 1845.
Sci Data 12, 525 (2025). https://doi.org/10.1038/s41597-025-04846-8Download citationReceived: 27 November 2024Accepted: 17 March 2025Published: 28 March 2025DOI: https://doi.org/10.1038/s41597-025-04846-8Share this articleAnyone you share the following link with will be able to read this content:Get shareable linkSorry, a shareable link is not currently available for this article.Copy to clipboard
Provided by the Springer Nature SharedIt content-sharing initiative