The protected flora of long‐established cemeteries in Hungary: Using historical maps in biodiversity conservation

Abstract The role of anthropogenically influenced habitats in conserving elements of the original wildlife has increased worldwide simultaneously with the disappearance of natural sites. Burial places are able to conserve original elements of the wildlife, and this fact has been known for at least a century. To this day, little is known about long‐time changes and the effect of long‐time management methods in cemeteries on the flora they harbor. The utility of historical maps in research focused on natural values, as well as in answering questions related to conservation was recently demonstrated, but the use of digitized historical maps in biodiversity research of the Carpathian Basin is very limited. In the present paper, we aimed to predict the conservation potential of long‐established and newly established cemeteries of Hungarian settlements with various population sizes based on the digitized maps of the 2nd Military Survey of the Austrian Empire (1819–1869), by categorizing cemeteries into 3 distinct (anthropogenic habitat, cemetery, or natural habitat) types. To build our models, we used records of the protected flora from Hungarian cemeteries, based on data of thematic botanical surveys of 991 cemeteries. Out of the surveyed cemeteries, 553 (56%) harbored protected plants, totaling 306.617 estimated individuals of 92 protected species, belonging to 28 plant families. These species represent 12% of the entire protected flora of Hungary. Hungarian cemeteries play a key role mainly in preserving steppe and dry grassland plant species. Long‐established and large cemeteries harbor more protected plant species than small and newly established ones. Human population size of the settlements correlated negatively with the number of protected species and individuals. Moreover, woodland cover and proportion of grassland also significantly positively affected the number of protected plant species in cemeteries.


| INTRODUC TI ON
The surface of Earth has changed drastically due to human activities (Fahrig, 2003). In this transformed landscape, man-made or human-influenced habitats serve as refuges for the original wildlife. Such habitats include road verges (Fekete et al., 2017(Fekete et al., , 2019, midfield islets (Cousins, 2006), and different burial sites, including burial mounds  or cemeteries, which often preserve the last remnants of natural habitat patches in the transformed urban and/or agricultural landscape. The appreciation of cemeteries in conserving biodiversity has seen an upsurge during the last few decades, and burial sites started to receive special attention in this respect. While studies of the natural values of cemeteries in Europe started with the examination or their bird communities (Pearson, 1915), by the end of the century, further studies were published, revealing the role of cemeteries in the conservation of bryophytes (Fudali, 2001;Mišíková & Kubinská, 2010), lichens (Hawksworth & McManus, 1989;Matwiejuk, 2009), dendroflora (Antkowiak & Heine, 2005), and vascular plants (Czarna, 2016;Šilc, 2009). Nonetheless, to date research on the role of cemeteries in wildlife conservation concentrated to only a few regions of the world (Barrett & Barrett, 2001;Löki, Deák, Lukács, & Molnár V., 2019). Furthermore, we lack thematic surveys that would cover multiple cemeteries in Europe. As such, published papers are geographically strongly biased, and the range of studied organisms is relatively narrow, even in the long-time-transformed, densely populated, and well-researched Europe.
Cultural, ethnographic, and though with less thoroughness but natural values of cemeteries were all deemed noteworthy in the past century in Hungary and in the Carpathian Basin by early works.
Nonetheless, despite the valuable historical studies that discuss both the natural and cultural values of Hungarian cemeteries, our knowledge on the role of cemeteries in nature conservation remained sporadic for a long time. In Hungary, ethnographic researchers started to pay greater attention to cemeteries in the second half of the 20th century. For instance, Balassa (1989: 149) suggested that mosaic land use of traditional Hungarian cemeteries, besides the long-term, traditionally stable, and sustainable management practices of burial places, has favored the conservation of natural habitat patches in the Carpathian Basin: "Hungarian peasant cemeteries usually have three sections. The first section, where no graves had been installed yet, are usually utilised as agricultural plots, where crops are grown and seeds are sown. The second part, usually the largest, where burials take place at the moment and where they took place during the last 35 years. The latter part slowly transits into the third part, where no graves actively cared-for can be found, and where bushes and trees grow without control, while grass starts to grow on increasingly flat graves. The oldest part of the cemetery is progressively being regained by the natural vegetation." Consequently, Seléndy (1972) also noted that cemeteries were green fields dominated by grassland vegetation ever since the Middle Ages, and were under protection by various sanctions against nature-damaging activities for a long time. Novák wrote in 1978 that "Anyone-just like in the case of forests or vineyards-who e.g. caused damage, cut wood, or allowed his/her animals to roam inside the cemetery… was obliged to pay a fine of 12 Hungarian Forints…, the delinquent shall be severely sanctioned, his cows or pigs grazing in the cemetery should be shot without warning." Cemeteries were regularly hand mown, and the rights to harvesting hay, fruits, and other goods from the cemetery were regularly auctioned until the end of the last century (Balassa, 1989). One of the cemeteries of the Aggtelek-Karst represents a good example: The successful bidder paid 1,4 EUR for the mowing and harvesting rights in 1977 (Kunt, 1978 (Margittai, 1915). He describes the endangerment of the sandy flora in Cséke and describes the surveyed cemetery as a refuge: "…the cantor's pasture taking up only a couple square yards… was not ploughed because the hillside is rather steep here… even on this soil of running sand a beautiful grassy pasture could have developed. The rare Pulsatilla grows on this small plot and the sandy flora of the surrounding area also found refuge here." Later in the North Hungarian Mountains, the flora of cemeteries was surveyed by Schmotzer (2015). He found many valuable plant species, including the endemic Thlaspi jankae in 23 cemeteries, a species listed in the Habitats Directive. Moreover, based on the recent field surveys in Hungary, the shrub Spiraea crenata L. previously thought to be extinct in Hungarywas shown to be present in several Hungarian cemeteries . Besides the latter species, 26 other valuable and protected plant taxa were identified during the surveys .
It has recently been revealed that historical maps can be used to detect long-term changes in the landscape. They represent useful tools in detecting change in land use (Cui et al., 2014;Skaloš et al., 2011;Skokanová et al., 2012) and vegetation (Kelly, Allen-Diaz, & Kobzina, 2005), in analyzing the structure of secondary forests (Mladenoff, White, Pastor, & Crow, 1993), or in determining water-level fluctuations of larger lakes (Zlinszky & Timár, 2013).
Historical maps are also appropriate tools for mapping ecosystem services, as well as their changes over a longer period of time (Jiang, Bullock, & Hooftman, 2013). Moreover, historical maps are effective tools in the reconstruction of certain sensitive, disappeared, or shrunken habitats in the landscape (Bromberg & Bertness, 2005).
They can also be of help in monitoring the loss and endangerment status of unique habitat types (species listed in the Habitats Directive, Biró, Bölöni, & Molnár, 2018). All in all, historical maps provide an exceptional source of information, providing a glimpse into past and an opportunity to collect data that can be utilized in active nature conservation today. As numerous digitized maps become The Eurasian forest-steppe represents one of the most diverse and complex ecosystems in the world, and it stretches from the Ural Mountains on the east to eastern Europe on the west (Erdős et al., 2018). Hungary is situated at the western edge of the Eurasian forest-steppe ecoregion (Fekete, Molnár, Kun, & Botta-Dukát, 2002), contributing to the high level of habitat complexity observed in the Pannonian Ecoregion. This unique region harbors many different habitat types on a relatively narrow spatial scale and holds a high level of species diversity (Erdős et al., 2014). Although the Carpathian Basin has an undisputable role in conserving dry grassland and forest-steppe habitats in Central Europe, most natural habitats have disappeared, and several species have become threatened in recent decades (Molnár, Bölöni, & Horváth, 2008).
Consequently, the importance of man-made habitats in conserving natural elements of the landscape has increased in the new era in parallel with the demand for studies in such unique ecoregions, like the westernmost part of the Eurasian forest-steppe ecoregion, that would survey these secondary habitats with the goal of identifying and preserving key conservation sites.
In this study, we aimed to carry out a systematic survey of protected plant species in cemeteries that has great conservation potential globally. During our study, we adopted a novel technique: Using historical maps, we aimed to test the predictability of conservation importance of specific burial sites. In detail, we aimed (a) to evaluate which protected plant species are present and how frequently do they occur in Hungarian cemeteries; (b) to test whether long-established cemeteries host more protected plants than newly established ones; and (c) to determine which are the key characteristics of cemeteries that render them suitable for protected plants.
We hypothesized that (1) Hungarian cemeteries host numerous protected plant species in their territory; (2) long-established cemeteries host more protected plant species than newly established ones; (3) cemetery size, and proportion of grassland and woodland cover will significantly affect the number of plant species and individuals in cemeteries. Our study represents the first systematic survey of the flora in cemeteries of Hungary and the first to implement the use of historic maps in evaluating botanical values of man-made habitats of the country.

| Field survey
We surveyed burial grounds between 2014 and 2017, regardless of their religious affiliation, spatial dimension, or geographic location within Hungary. Surveyed cemeteries were selected without any previous knowledge on their history or flora. A total of 991 by a cemeteries were surveyed, including cemeteries from every county of Hungary.
Although the size of cemeteries was not uniform, we endeavored to a complete survey of the sampled cemetery's area; if the area of the cemetery was more than 5 hectares, we focused our searches to reserve areas of cemeteries along the fence. All protected plant taxa were identified and recorded, and the number of individuals belonging to each of these was counted or in the case of more than 500 individuals, estimated in each cemetery. Taxa

| Historical maps
In order to identify the age of the surveyed cemeteries, we down- that could not be found on the historic maps were regarded as newly established cemeteries. It is also highly presumable that every settlement had its own burial sites, but as historical maps not always depicted cemeteries at every settlement, we conclude that the lack of cemetery in the close proximity of a given settlement rather means that military surveys considered the smallest cemeteries insignificant for military interests. Therefore, the smallest ones were not indicated on maps; it also follows that if a cemetery was small enough to miss out from the used historical map, then we categorized it to our analysis as newly established, as it probably had a small potential for conserving significant biodiversity in its area. Furthermore, we measured the total area, proportional coverage of woodland vegetation, grassland, graves, and the distance of cemeteries from settlement margins for each surveyed cemetery based on recent satellite images retrieved from the Google Earth Pro software.  (Table   S1).

| The age of cemeteries and their conservation potential
The historical habitat type of the cemeteries significantly affected the number of species they currently harbor. Long-established cemeteries hosted more protected plant individuals than newly established ones (categorized as other habitat types on the historical maps, Figure 4.), while no correlation was found between historical habitat type and the number of protected plant individuals present.

| Factors affecting species richness and abundance
The ZIP and ZINB models built to explain the number of protected  (Table 3.).

| D ISCUSS I ON
The current paper presents the first thematic field survey of the pro- Our results also highlight that the conservation potential of cemeteries varies widely across different regions of the country.
This pattern can be attributed to both biogeographic patterns and land-use factors. Similarly to most places in the world, at the time of the establishment the appropriate place for the cemeteries in Hungary was selected based on practical, for example, security and religious considerations (Seléndy, 1972:19), while esthetics was only a secondary criterion. The dynamic mosaic structure of traditional cemeteries mentioned by Balassa (1989)

TA B L E 2
Factors explaining variation in the number of protected plant species in Hungarian cemeteries, as inferred from zero-inflated Poisson generalized linear models part of nature, but people could moderately alter the cemetery for both burial and extensive economic purposes (e.g., orchards and hay meadows), while in parallel with these, excessive nature destructive actions were strictly forbidden in most cemeteries (Novák, 1978).
Historical data retrieved from old maps indicated that the number of habitats of a given territory can decrease significantly over time (Jiang et al., 2013). Another study that used historical maps to reconstruct past conditions showed that the semi-natural grassland cover decreased by over 96% in the studied area since the beginning of the 20th century in Sweden (Cousins, Auffret, Lindgren, & Tränk, 2015). An excellent study has been conducted about habitat loss in Hungary by evaluating 5,000 randomly selected localities over 7 periods from 1783 to 2013 (Biró et al., 2018): According to this study, trends in habitat loss over time are strongly habitat-specific, and most habitats had continuous loss from period to period. As a consequence, cemeteries are currently emerging as small islands that conserve natural and semi-natural grasslands in the radically changed landscape, as indicated by the high number of protected grassland and steppic plant species detected in Hungarian cemeteries. However, it must be added that they are not exclusively the guardians of grassland and steppic organisms in the close proximity of urban areas: According to the findings of Deák, Valkó et al. (2020), burial mounds (kurgans) also play a key role in conserving at least grassland specialist plants and arthropod species, while their potential in conserving other organisms is also presumably high.
Our results indicate that long-established cemeteries harbored significantly more protected plant species than newly established ones. This result supports the hypothesis that cemeteries represent habitat patches that more or less preserved their natural states, and provide safe harbors for natural habitats (and their protected organisms). These habitats remained only slightly disturbed over time and provided colonization windows for rare and sensitive organisms, and also provided a kind of micromosaicism in the habitat. It is very important to note that long-established cemeteries did not host more protected plant individuals than newly established ones. We believe that this lack of association might indicate the effect of incorrect management practices. Although old cemeteries have higher conservation potential, due to their higher biodiversity, especially with respect to protected species, their incorrect management might keep these species at bay and prevent their population growth. On the other hand, cemetery area has been shown to positively correlate with the number of protected individuals they harbor. Consequently, the species composition F I G U R E 5 Long-established cemeteries host more protected plant species than those converted from anthropogenic habitats into cemeteries relatively recently (i.e., after the 2nd Military Survey)

TA B L E 3 Factors explaining variation
in the number of individuals of protected plant species in Hungarian cemeteries, as inferred from zero-inflated negative binomial generalized linear models appears to be more representative of the conservation potential of a cemetery, than the number of individuals belonging to protected taxa found in them. This pattern might emerge due to the fact that many years of improper management is necessary to entirely eliminate a species from a population, but a high proportion of individuals can be eradicated with only a few years of intensive management, especially in the case of grasslands (e.g., Plantureux, Peeters, & McCracken, 2005).
We found that more protected plant species can be found in cemeteries at higher altitudes (Figure 1). This pattern might emerge from the different species pools of different climatic zones at different altitudes. However, according to our personal observations, radical transformation of the landscape, frequent use of herbicides, and use of motorized mowing devices are more frequent in lowland cemeteries, rendering these less suitable for rare and sensitive organisms. Moreover, lowland areas contain more settlements with higher number of inhabitants, which can explain the more intensive management adopted in lowland cemeteries and the concentration of rare organisms at higher altitudes. Therefore, we suggest that the greatest attention of both professional conservation biologists and local people should be devoted to cemeteries of small villages. Given that most sites harboring species of conservation concerns are exempt from nationwide level protection, we press that these cemeteries should urgently receive protection at least on the regional level. Our observations are in consent with previous field surveys conducted in Turkey , where one of the key factors explaining the conservation value of cemeteries was the type (village, small town, or city) of the nearby human settlement.
The latter study suggested that urbanization is a key threatening factor to natural values in cemeteries. We believe that it is a pressing need to identify newly adopted management practices of cemeteries that are unsustainable for the conservation of natural values on the long term.
In parallel with the change in traditional management practices, the structure of the flora has also changed in Hungarian cemeteries over the past decades ( Figure 4) in Hungarian cemeteries, often as widespread species that radically transform the landscape of burial places (Balassa, 1989).
The religious affiliation of cemeteries is considered to be a key factor affecting the landscape and wildlife diversity in their territory (see . According to ethnographic researchers, Calvinist cemeteries are less cared compared with Catholic ones (Balassa, 1989), also reflected by the lower frequency and number of ornamental plants in the former. Accordingly, some ethnographers concluded that Calvinist cemeteries are more neglected (Balassa, 1989:149), while detailed analyses need to be applied to compare the effects of religious affiliation on natural values in more European cemeteries.
Novel cemetery management practices are highly divergent from traditional ones, and a "well-maintained" cemetery implies intensified pressure on the natural vegetation (e.g., motorized mowing, herbicides, use of concrete). Modern gardeners usually favor low-mowed lawns. For example, a Hungarian cemetery managing handbook recommends that if the grass reaches 8 cm in height the area should be mowed (Seléndy, 1972:224). Nonetheless, Hungarian ethnographers, historians, suggest that the extremely low, monodominant vegetation is unanimous, just like the modern cemetery landscape, dominated by high density, large concrete graves (Pocsai & Tóth, 1990:8). We suggest that hand mowing would be the most appropriate method in order to preserve these grassland habitats Afforested cemeteries are well-functioning, while also representing valuable and esthetic parks, providing potential recreational benefits for people, especially in the urban landscape. However, such cemeteries cannot be realized on the long run, if the modern burial protocols continue to be adopted and implemented with intensities seen during the last century. "We shall not look for immortality in cold and lifeless stones" (Pocsai & Tóth, 1990:10). Instead, we should adopt new and progressive ways to rejuvenate the area of cemeteries. For instance, according to Holden and McDonald-Madden (2018) conservation burials are carried out in the highest numbers in the UK, the United States, and Canada. The fees paid for these burials are used to fund the acquisition, protection, restoration, and management of lands that benefits human and environmental well-being.
We suggest that further studies should be conducted to identify the role of cultural factors and religious affiliations in determining the conservation value of cemeteries. Our results also suggest that thematic surveys of cemeteries that concentrate on the detection of multiple taxa, especially on a wide geographic scale, would provide valuable information on the role of anthropogenically influenced habitats in conserving valuable habitats and sensitive or rare organisms. Based on this systematic survey, we conclude that historical maps provide an excellent opportunity in biodiversity conservation to effectively designate sampling areas with presumably high conservation value for further research.

CO N FLI C T O F I NTE R E S T
None declared.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data used in the analyses are available from Dryad-https://doi. org/10.5061/dryad.jwstq jq6c (Löki et al., 2020).