Forest Ecology and Management 330 (2014) 228–239

Contents lists available at ScienceDirect

Forest Ecology and Management

journal homepage: www.elsevier .com/ locate/ foreco

Determining ancient woodland indicator plants for practical use:A new approach developed in northwest Germany

http://dx.doi.org/10.1016/j.foreco.2014.06.0430378-1127/� 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding author. Tel.: +49 551 69 401 313; fax: +49 551 69 401 160.E-mail address: moelder@gmx.de (A. Mölder).

Marcus Schmidt a, Andreas Mölder a,⇑, Egbert Schönfelder a, Falko Engel a, Inga Schmiedel b,Heike Culmsee c

a Northwest German Forest Research Station, Department A (Forest Growth), Section Forest Conservation and Natural Forest Research, Grätzelstraße 2, D-37079 Göttingen, Germanyb Georg-August University Göttingen, Albrecht von Haller Institute for Plant Sciences, Department Vegetation and Phytodiversity Analysis, Untere Karspüle 2, D-37073 Göttingen,Germanyc DBU Natural Heritage, German Federal Foundation for the Environment, An der Bornau 2, D-49090 Osnabrück, Germany

a r t i c l e i n f o a b s t r a c t

Article history:Received 24 April 2014Received in revised form 26 June 2014Accepted 27 June 2014

Keywords:Floristic datasetsForest speciesHistorical mapsNature conservationRecent woodlandHabitat continuity

Ancient woodlands that have been in continuous existence for hundreds of years have a floristiccomposition which greatly differs from younger afforestations. The occurrence of certain associatedvascular plant species, termed ‘‘ancient woodland indicator plants‘‘, can be used to recognise the continuityof woodland cover. Ancient woodland habitats frequently contain a typical and rich forest biodiversityand can often be regarded as ‘‘biodiversity hotspots’’. To pinpoint these habitats for nature conservation,there is a need to compile ancient woodland indicator lists with a widespread validity.

In this study, we introduce a new methodical approach that enables the compilation of such lists fromthe readily available resources of plant species monitoring programs, archive records, and land coverdata. Using northwest Germany as a model region, we have developed an ecologically grounded list of67 ancient woodland indicator plants for this area. In this context, we consider the ‘‘ancient woodlandindicator plants’’ as a subset of the larger group of ‘‘ancient woodland plants’’.

The widely applicable ancient woodland indicator plants list presented here may be a useful tool forfuture forest nature conservation. Potential applications include: (a) the identification of ancientwoodlands in areas where historical maps are lacking, (b) the identification of biodiversity hotspots ofancient woodland indicator plants, and (c) locating ancient semi-natural woodlands.

Finally, we highlight the importance of effective conservation management, which should seek topromote the typical plant diversity of ancient semi-natural woodlands. In doing so, conservationmanagement should promote the preservation of remaining ancient deciduous woodlands and inhibitthe conversion of ancient woodlands to coniferous or mixed forests.

Additionally, conservation management should strengthen the connections between recent and ancientwoodlands through habitat corridors. Furthermore, careful forest management of deciduous ancientwoodland sites with high typical woodland plant diversity has to be ensured to avoid soil damage.

� 2014 Elsevier B.V. All rights reserved.

1. Introduction

The continuity of woodland cover in time is regarded as a keyfactor for biodiversity in temperate forest ecosystems (Peterken,1974; Singleton et al., 2001; Hermy and Verheyen, 2007; Moningand Müller, 2009; Nascimbene et al., 2013). Numerous studieshave shown that woodlands in existence for hundreds of yearsdiffered greatly from younger afforestations with regard to theirfloristic composition (Wulf, 2003; Ito et al., 2004; Hermy and

Verheyen, 2007; Svenning et al., 2008; Kelemen et al., 2014). Thisdiscrepancy is particularly distinctive in regions with a lowproportion of woodland cover and a high degree of fragmentation(Ferris and Humphrey, 1999; Hermy et al., 1999; Wulf, 2003). Incontrast, the linkage between woodland continuity and theoccurrence patterns of woodland plant species is lower in areaswhere the majority of woodland is ancient and features a smallerdegree of ecological isolation (Dzwonko and Gawroński, 1994;Ferris and Humphrey, 1999; Schmidt et al., 2009).

In Great Britain, the term ‘ancient woodland’ defines land thathas been continuously wooded since at least 1600 AD (Spencerand Kirby, 1992; Goldberg et al., 2007; Stone and Williamson,2013). In our study, with a focus on the highly fragmented ancient

http://crossmark.crossref.org/dialog/?doi=10.1016/j.foreco.2014.06.043&domain=pdfhttp://dx.doi.org/10.1016/j.foreco.2014.06.043mailto:moelder@gmx.dehttp://dx.doi.org/10.1016/j.foreco.2014.06.043http://www.sciencedirect.com/science/journal/03781127http://www.elsevier.com/locate/foreco

M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239 229

woodland of northwest Germany, we refer to ‘ancient woodland’ asland that has been continuously wooded since at least 1800 AD,since only from this point on are area-wide coverage data onhistorically old woodland sites available (Wulf, 2003; Glaser andHauke, 2004). Ancient woodlands with a long habitat continuityharbour a high number of rare and threatened species, and aretherefore of particular importance for nature conservation (Rose,1999; Rackham, 2003; Hermy and Verheyen, 2007; Leuschneret al., 2014).

Based on 22 regional studies from northwest and CentralEurope, Hermy et al. (1999) compiled a list of 132 vascular plantsclosely linked to ancient woodland. Verheyen et al. (2003)evaluated 20 field studies from eight European countries and fournortheast American states that compared the vegetation of ancientand recent forests. From these they concluded that the response offorest plant species to land use coincided with a clustering ofspecies featuring different ecological characteristics. In regard tothis, slowly colonising species, many of which occur in ancientforests, are typically characterised by low dispersibility(Verheyen et al., 2003; Kimberley et al., 2013).

The continuity of woodland cover can be recognised by meansof certain associated vascular plant species, known as ‘‘ancientwoodland indicator plants’’ (Rose, 1999; Glaves et al., 2009). InGreat Britain, several regionalised lists of ancient woodlandindicator plants have been compiled over the last 30 years (Rose,1999; Glaves et al., 2009), initiated by the work of Peterken (1974).In this study, we consider the ‘‘ancient woodland indicator plants’’as a subset of the larger group of ‘‘ancient woodland vascularplants’’, even though there is no clear differentiation in literature.

Ancient woodland habitats frequently contain a typical and richforest biodiversity and can often be regarded as ‘‘biodiversityhotspots’’. Ancient woodland indicators are an important tool todetermine these valuable habitats (Myers et al., 2000; Hermyand Verheyen, 2007; Meyer et al., 2009; Mölder et al., 2014a).Furthermore, the occurrence of ancient woodland indicator plantscan be an indicator for the conservation value of adjacent openareas. In this regard, Diekmann et al. (2008) pointed out that forestand open-habitat specialists respond not only similarly tolandscape heterogeneity and environmental gradients, but also toregional patterns of land use and habitat continuity.

According to Hermy et al. (1999), due to the distinct localvariation in the ecological behaviour of forest plant species, regionallists of ancient woodland indicator plants are more appropriatethan one Pan-European list. This point of view has been supportedby numerous other authors (e.g., Rose, 1999; Wulf, 2004; Glaveset al., 2009; Perrin and Daly, 2010). However, for the applicationof ancient woodland indicator lists (e.g., by nature conservationauthorities or woodland surveyors), it is more convenient to coverlarger areas at the supra-regional greater landscape level inorder to achieve enhanced validity and comparability. Here, wepresent a new methodological approach for the identification ofsupra-regionally implementable ancient woodland indicatorplants. In contrast to previous studies, we have not adopted ourancient woodland indicator plant list from a number of previoussingle studies or local observations. Instead, we systematicallyevaluated plant distribution data of floristic surveys in relation toancient woodland cover data from state-wide inventories. In doingso, we determined ancient woodland indicators using consistentand repeatable statistical methods. We have used the large areaof northwest Germany as a model region. Here, in these Pleistocenelowlands, ancient woodlands are scattered and their extent isrelatively low (Glaser and Hauke, 2004). We would therefore expecta strong association of certain woodland plant species with thesewoodlands. In addition, the study area is covered by a mappingprogram of the distribution of vascular plants with a resolutionof ca. 30 km2 and so provides a promisingly large data set.

In developing the ancient woodland indicator plant method, weaddressed the following questions:

(1) Which forest plant species can be classified as supra-regionally valid ancient woodland plants for the area ofnorthwest Germany?

(2) Are there groups of ancient woodland plants that are relatedto certain environmental conditions of different woodlandtypes?

(3) Which of the ancient woodland plants are suitable indicatorsfor application in forestry and nature conservation practice?

2. Materials and Methods

2.1. Study area

The study was conducted in northwest Germany and coveredthe entire federal states of Schleswig–Holstein and Bremen andthe lowland parts of the state of Lower Saxony (altogether coveringa total area of 53,549 km2). We delimitated the borderline betweenthe lowland and the upland parts of Lower Saxony by followingGarve et al. (2007). Based on the German network of topographicalmaps (scale 1:25,000), the study area was divided into a gridof 2378 quadrants, of which each grid cell had a resolution ofapproximately 5.5 � 5.5 km or 30 km2 (Fig. 1).

In the Pleistocene lowlands of northwest Germany, naturalwoodlands would be dominated by deciduous tree species,especially beech (Fagus sylvatica). However, as elsewhere in CentralEurope, there are no remaining woodlands completely unaffectedby long-term human activity (Szabó, 2009; Ellenberg andLeuschner, 2010; Arnold, 2011). The middle of the 18th centurysaw initial attempts to establish conifer plantations on infertileheathlands; a century later, for the first time, coniferous and mixedforests (consisting of broadleaved and coniferous trees) reachedsignificant proportions (Niemann, 1809; Kremser, 1990; Hase,1997). Since then, even deciduous stands on ancient woodlandsites have been converted to conifer plantations or mixed forests(see Table 1; ‘‘coniferous ancient woodland’’ or ’’mixed ancientwoodland’’). This is especially true for nutrient-poor sites (Glaserand Hauke, 2004). Currently, 26% of the woodlands in our studyarea are ancient. The proportion of deciduous ancient woodlandamounts to 7%. In contrast to other European regions (e.g., partsof Great Britain), coppicing played only a minor role in northwestGerman ancient woodlands during the last 200 years (Kremser,1990; Hase, 1997; Rackham, 2003).

2.2. Data sets

The floristic data for Lower Saxony and Bremen were obtainedfrom the database of the Lower Saxon plant species monitoringprogram (NLWKN 1982–2003; Garve et al., 2007). For Schleswig–Holstein, floristic data was collected by Raabe (1987) and the AGGeobotanik (2013) from 1961 to 2012. From these data sets, weconsidered the 452 vascular plant species that are closely boundto forest habitats according to the German Forest Vascular PlantSpecies List (Schmidt et al., 2011). 164 species belong to category1.1 (largely restricted to closed forests), 38 species to category1.2 (preferring forest edges and clearings), and 250 species tocategory 2.1 (occurring in forests, as well as in open habitats).For each of these plant species, we ascertained the occurrence(presence or absence) in each topographic map quadrant.Nomenclature followed Wisskirchen and Haeupler (1998).

We determined the ancient woodland area (area_aw) andproportion (perc_aw) in each quadrant, distinguishing respectivelybetween ancient woodland sites currently dominated bydeciduous tree species (perc_daw), coniferous tree species

Fig. 1. The study area of northwest Germany, including the entire states of Bremen and Schleswig–Holstein and the lowlands of the state of Lower Saxony. The displayed2378 grid cells are the basic units for the vascular plant survey programs.

Table 1Woodland area and proportions of recent and ancient woodlands in the study area of northwest Germany.

Woodland type Variables Area (ha) Proportion (%)

Total woodland area 829,252 100Recent woodland (younger than ca. 200 years) 612,817 73.9Ancient woodland (older than ca. 200 years) aw 216,435 26.1

Deciduous ancient woodland daw 61,657 7.4Mixed ancient woodland maw 67,915 8.2Coniferous ancient woodland caw 86,863 10.5

230 M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239

(perc_caw) and a mixture of both types (perc_maw) (Table 1). ForSchleswig–Holstein, data on ancient woodland was obtained fromGlaser and Hauke (2004), and for Lower Saxony and Bremen weused high-resolution data provided by the Lower Saxon forestplanning agency. Both data sources utilised historical land surveymaps (compiled mostly between 1750 and 1800) and youngertopographical maps in order to determine whether current woodlandhas been continuously wooded since 1800 or not. Woodland withforest continuity since at least 1800 was regarded as ancient (Wulf,2003; Glaser and Hauke, 2004), and information on current treespecies composition has been derived from forest inventories.

All spatial data was processed in QGIS (v. 2.2; QGISDevelopment Team, 2014). We removed 671 quadrants (grid cells)without forest cover and/or plant species occurrence from the data

set, thus the combined data on plant species and woodlanddistributions for 1707 quadrants were used in the final analysis.

2.3. Statistical analysis

2.3.1. Identification of supra-regional ancient woodland plantsBased on a sequential matrix (M1, in which each row describes

the occurrence of a species in an arbitrary quadrant), we computedan incidence matrix (M2) as follows:

M2 ¼ ½mi;j�; ð1Þ

where i is equal to quadrants 1–1707, and j is equal to species1–452, with mi,j either having the value 0 or 1 (binary values).

M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239 231

Another matrix M3, contained data on thefive ancient woodland variables for each quadrant(perc_aw, perc_daw, perc_maw, perc_caw, area_aw; Table 1),either expressed in hectares (_area) or as a percentage (_perc).Both matrices M2 and M3 were joined, on the basis of the uniquenumber of the quadrants, to generate matrix M4. Based on matrixM4, for each combination [species * ancient woodland variable] ageneralised linear model (GLM) for binary data (Fahrmeir et al.,2009) was computed. If a species was present in less than 0.5% ofall quadrants, it was excluded from further analyses due topossible convergence difficulties. Finally, 390 of the 452 speciesremained in the analysis. The most frequent species Glechomahederacea occurred in 99.0% of all investigated quadrants.Therefore, no upper threshold was necessary. The resulting teststatistics for the regression parameter were used for creatingmatrix M5 (displayed in a variance table, Appendix Table A.1),in which pi,j corresponded to the test statistics zi,j of each GLM:

M5 ¼ ½pi;j�; ð2Þ

where i is equal to species 1–390, and j is equal to ancient woodlandvariables 1–5.

Since all the GLMs featured the same sample size (number ofquadrants) and the same structure, we were able to interpret thez-values without further weighting: with regard to an ancientwoodland variable, z-values around 0 indicated no relationship,while high z-values (>10) indicated a very close connectivity;negative z-values indicated a connectivity to recent woodland.

This variance table was furthermore used for conducting aprincipal component analysis (PCA; cf. Venables and Ripley,2002). In the PCA, we considered five ancient woodland variables(perc_aw, perc_daw, perc_maw, perc_caw, area_aw; Table 1). Thevariable ‘‘area of ancient woodlands’’ (area_aw) in addition to thevariable ‘‘proportion of ancient woodlands’’ (perc_aw) providedadditional information on the significance of forest area for thedistribution patterns of plant species.

A biplot was created, which allowed for the analysis not only ofthe correlation between the variables, but also of the relationshipbetween the ancient woodland variables and the species.

By the use of k-means clustering (cf. Venables and Ripley, 2002),we grouped all species into seven clusters, which were interpretablein a meaningful way. The number of seven clusters was confirmedby applying the R software with the ‘‘clValid’’ package (Brock et al.,2008). A combined presentation (biplot) of the clusters and the(species) coordinates of the first and second PCA axis allowed forthe interpretation of relationships between cluster composition,species occurrence and ancient woodland variables. With thepurpose of interpreting the seven clusters ecologically in ouranalysis, we also included Ellenberg indicator values (EIV) for light,reaction, nitrogen and moisture (Ellenberg et al., 2001). EIV andz-values were tested for differences between the seven speciesclusters (Kruskal Wallis H-test, p 6 0.05, with subsequentBonferroni-corrected Wilcoxon rank-sum test). For the few speciesthat were lacking particular EIV, we calculated auxiliary indicatorvalues by averaging over all quadrants. In order to then fit theEIV onto the PCA plot, we used the function ‘‘envfit’’ provided bythe ‘‘vegan’’ package in R (Oksanen et al., 2012).

All statistical analyses were performed by using the R softwareversion 3.0.1 (R Development Core Team, 2013) with the ‘‘vegan’’package (Oksanen et al., 2012) and the ‘‘clValid’’ package (Brocket al., 2008). Significance of statistical tests was noted as follows:*** = p 6 0.001; ** = p 6 0.01; * = p 6 0.05; n.s. = p > 0.05.

2.3.2. Compilation of an ancient woodland indicator plant list forpractical use

The GLM, PCA and cluster analyses served as a screeningprocedure. In order to derive an ancient woodland indicator plant

list from the previously compiled variance table, we applied thosespecies groups found to be indicative of ancient woodlands to anindependent dataset. This dataset consisted of point data recordsfor vascular plants (1980–2013; AG Geobotanik, 2013) and adetailed ancient woodland inventory of the Schleswig–HolsteinState Forests (Dubberke-Spandlowski, 2011). The Schleswig–HolsteinState Forests comprise 50.000 ha of woodland scattered allover Schleswig–Holstein.

For all plant species of the aforementioned dataset reported inmore than 25 points (with the omission of extreme rarities; seeRose, 1999), we calculated the percentage of point data recordssituated on ancient woodland sites. In doing so, regions with a highdensity of ancient woodland indicators could be separated fromthose hosting none or only few of these species. Non-native treespecies were also excluded from this analysis. If P75% of a species’occurrences were situated in ancient woodlands, that particularspecies was then regarded as an ancient woodland indicator plant.The 75% threshold was chosen according to the interpretation of theRevised Index of Ecological Continuity (Coppins and Coppins, 2002).

3. Results

3.1. Identification of supra-regional ancient woodland plants

Variance analysis resulted in z-values of 390 species independence of five ancient woodland variables (AppendixTable A.1). Z-values ranged from 19.2 to �16.8 (Fig. 2). Theproportion of ancient woodland (perc_aw) was strongly determinedby the proportion of deciduous woodland on ancient woodlandsites (perc_daw), as shown by the very close correlation of thez-values of both variables (Pearson’s r = 0.99, p-value 6 0.001). Thisrelationship was also obvious from the results of the PCA (Fig. 3).Similarly, the proportion of coniferous forests on ancient woodlandsites (perc_caw) and the proportion of mixed forests on ancientwoodland sites (perc_maw) were almost congruent in PCA resultsand correlation of the z-values (r = 0.84, p-value 6 0.001).

As a result of the k-means cluster analysis, the list of 390 specieswas divided into seven groups (Table 2), which were each namedafter a typical plant species. The seven groups were ordered accordingto their preference for ancient or recent woodlands and to Ellenbergindicator values (EIV). With respect to woodland continuity, threegroups of ancient woodland plants (A, Galium odoratum group;B, Mercurialis perennis group; C, Oxalis acetosella group) weredistinguished from one group of recent woodland plants (G, Agrostiscapillaris group), and three further groups of more or less indifferentspecies (D, Ranunculus ficaria group; E, G. hederacea group;F, Deschampsia flexuosa group) (Fig. 4). These groups largely differedin their z-values (Fig. 2) and in their Ellenberg light values (Fig. 4,Table 2, correlation with the first PCA axis: r = 0.33, p-value6 0.001). The groups showed a gradation, from the G. odoratum groupindicating darkest conditions to the A. capillaris group indicatinglightest forest floor conditions. Furthermore, the groups showedlargely varying reaction values (Fig. 4, Table 2, correlation withthe second PCA axis: r = 0.32, p-value 6 0.001), with the M. perennis,G. odoratum, and R. ficaria groups indicating most base-rich soilconditions. Across all groups, Ellenberg nitrogen values exerted aminor influence and soil moisture values were not significant.

Considering z-values (Fig. 2), the G. odoratum group (A in Fig. 4,Appendix Table A1) was most closely associated with ancientdeciduous woodland. This group was mainly characterised byshade-tolerant plant species. Most of these are indicators ofmoderately acidic to weakly basic soils (Table 2). Species of theM. perennis group (B in Fig. 4) were also strongly connected todeciduous ancient woodlands (Fig. 2), but the species of thisgroup, compared to the G. odoratum group, occurred on woodlandsites with better conditions of light and base supply (Table 2). In

Table 2Species numbers, mean z values of the variable ‘‘proportion of deciduous woodland on ancient woodland sites’’ (perc_daw), and Ellenberg indicator values (EIV) of the seven species groups. p values of significant differences between theEIV of two groups are given in bold. *** = p 6 0.001; ** = p 6 0.01; * = p 6 0.05; n.s. = p > 0.05. SD = standard deviation.

Ancient woodland plants Indifferent woodland plants Recent woodland plants

Groups Galium odoratum Mercurialis perennis Oxalis acetosella Ranunculus ficaria Glechoma hederacea Deschampsia flexuosa Agrostis capillaris

Species number 45 48 70 102 57 49 19Mean z values (perc_daw) 14.8 12.6 7.6 4.6 �1.8 0.2 �11.0z values (perc_daw), SD 2.2 2.5 1.8 1.6 2.4 1.9 2.9

Mean EIV for light 4.3 5.2 4.8 5.9 6.3 6.2 6.3EIV for light, SD 1.6 1.7 1.4 1.5 1.1 1.3 1.2

Groups Galium odoratum – – – – – – –Mercurialis perennis n.s. – – – – – –Oxalis acetosella n.s. n.s. – – – – –Ranunculus ficaria 60.001*** n.s. 60.001*** – – – –Glechoma hederacea 60.001*** n.s. 60.001*** n.s. – – –Deschampsia flexuosa 60.001*** 0.03* 60.001*** n.s. n.s. – –Agrostis capillaris 0.002** n.s. 0.008** n.s. n.s. n.s. –

Mean EIV for reaction 6.3 7.0 5.1 6.2 5.6 4.3 4.3EIV for reaction, SD 1.2 1.0 1.7 1.8 2.1 1.8 1.5

Groups Galium odoratum – – – – – – –Mercurialis perennis n.s. – – – – – –Oxalis acetosella 0.01** 60.001*** – – – – –Ranunculus ficaria n.s. n.s. 0.01** – – – –Glechoma hederacea 60.001*** 60.001*** n.s. 60.001*** – – –Deschampsia flexuosa n.s. n.s. n.s. n.s. n.s. n.s. –Agrostis capillaris 0.02* 60.001*** n.s. n.s. n.s. n.s. –

Mean EIV for nitrogen 5.7 5.8 5.0 4.6 4.6 3.7 5.8EIV for nitrogen, SD 1.4 1.6 1.9 2.0 2.5 2.1 2.2

Groups Galium odoratum – – – – – – –Mercurialis perennis – – – – – – –Oxalis acetosella n.s. n.s. – – – – –Ranunculus ficaria n.s. 0.03* n.s. – – – –Glechoma hederacea 60.001*** 60.001*** 0.03* n.s. – – –Deschampsia flexuosa n.s. n.s. n.s. n.s. n.s. – –Agrostis capillaris n.s. n.s. n.s. n.s. n.s. n.s. –

Mean EIV for moisture 6.1 5.7 5.8 6.5 6.2 5.4 4.8EIV for moisture, SD 1.2 1.3 1.5 2.2 2.1 1.9 0.9

Groups Agrostis capillaris 0.03* n.s. n.s. n.s. n.s. n.s. –All other groups n.s. n.s. n.s. n.s. n.s. n.s. n.s.

232M

.Schmidt

etal./Forest

Ecologyand

Managem

ent330

(2014)228–

239

Fig. 2. Variation in z-values of the variable ‘‘proportion of deciduous woodland onancient woodland sites’’ (perc_daw) given for different species groups: A – Galiumodoratum group, B – Mercurialis perennis group, C – Oxalis acetosella group,D – Ranunculus ficaria group, E – Glechoma hederacea group, F – Deschampsiaflexuosa group, G – Agrostis capillaris group. Significant differences are indicated bydifferent lower case letters.

M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239 233

contrast, species of the O. acetosella group (C in Fig. 4) differed fromthe aforementioned groups concerning their ecological behaviourand in their degree of linkage to ancient woodlands (Fig. 2). Only

Fig. 3. PCA/biplot of the data listed in the variance table (Appendix Table A1). Matrix: 39axes 1 and 2 = 0.93). Abbreviations of the species names: see Appendix Table A1perc_aw = proportion of ancient woodlands in the total forest area per quadrant (%), percarea per quadrant (%), perc_caw = proportion of coniferous forests on ancient woodland son ancient woodland sites in the total forest area per quadrant (%), area_aw = area of an

a few species in this group reached high z-values (>10). Overall,species of this group were less strongly linked to ancient deciduouswoodlands (Fig. 2) and also tended to occur in coniferous or mixedancient woodlands (Figs. 3 and 4). The respective species oftenpreferred medium shade and grew on acidic to weakly acidic soils(Table 2).

The R. ficaria group (D in Fig. 4) was largely indifferent towoodland continuity. Nevertheless, several species of this groupwere characterised by comparatively high z-values in relation tothe proportion of ancient woodland and the proportion of deciduouswoodland on ancient woodland sites (Fig. 2). This showed that inthe group definition it was difficult to draw absolute limits. Thespecies of the R. ficaria group preferred semi-shady conditionsand grew on moderate acidic to basic soils (Table 3). The speciesof the D. flexuosa group, as well as the G. hederacea group, wereboth indifferent concerning habitat continuity, or even showedhigher affinity to recent woodlands. The former group (F inFig. 4) included medium shade tolerant plants or plants occurringin light shade, and indicators for acidic up to moderate acidic soils(Table 2). Species of the latter group (E in Fig. 4) preferred mediumshade conditions and moderately to weakly acidic soils (Table 3).Finally, the A. capillaris group included recent woodland species(G in Fig. 4), which mostly grew in locations receiving a highamount of light and preferred acidic or moderately acidic soils(Table 3). The species of this group were characterised by negativez-values in relation to the proportions of ancient woodland anddeciduous woodland on ancient woodland sites (Fig. 2).

We found obvious differences in the linkage to forest habitatsbetween the seven plant species groups (Fig. 5). The proportionof species that were largely restricted to closed forests (category

0 forest species (axis 1: eigenvalue = 3.31, axis 2: eigenvalue = 1.32, combined R2 of. For reasons of clarity most species names have been replaced by asterisks._daw = proportion of deciduous forests on ancient woodland sites in the total forestites in the total forest area per quadrant (%), perc_maw = proportion of mixed forestscient woodlands in the total forest area per quadrant (ha).

Fig. 4. PCA of the 390 forest species listed in the variance table (AppendixTable A1). The position of the species corresponds to Fig. 3, the letters indicate the 7groups identified by k-means clustering: A = Galium odoratum group (45 species),B = Mercurialis perennis group (48 species), C = Oxalis acetosella group (70 species),D = Ranunculus ficaria group (102 species), E = Glechoma hederacea group (57species), F = Deschampsia flexuosa group (49 species), G = Agrostis capillarisgroup (19 species). R = Ellenberg reaction value, N = Ellenberg nitrogen value,L = Ellenberg light value, M = Ellenberg moisture value.

Table 3List of ancient woodland indicator plants for northwest Germany. FSG = Forest speciesgroup according to the German Forest Vascular Plant Species List (Schmidt et al.,2011), 1.1 = largely restricted to closed forests, 1.2 = preferring forest edges andclearings, 2.1 = occurring in forests, as well as in open habitats.

No. Species name FSG Woodland species group

1 Actaea spicata 1.1 Mercurialis perennis group2 Allium ursinum 1.1 Oxalis acetosella group3 Anemone ranunculoides 1.1 Mercurialis perennis group4 Arum maculatum 1.1 Galium odoratum group5 Blechnum spicant 1.1 Oxalis acetosella group6 Brachypodium sylvaticum 1.1 Galium odoratum group7 Campanula trachelium 1.1 Mercurialis perennis group8 Cardamine bulbifera 1.1 Mercurialis perennis group9 Carex pallescens 2.1 Galium odoratum group

10 Carex remota 1.1 Galium odoratum group11 Carex strigosa 1.1 Mercurialis perennis group12 Carex sylvatica 1.1 Galium odoratum group13 Carpinus betulus 1.1 Oxalis acetosella group14 Chrysosplenium alternifolium 1.1 Galium odoratum group15 Chrysosplenium oppositifolium 1.1 Galium odoratum group16 Circaea alpina 1.1 Oxalis acetosella group17 Circaea lutetiana 1.1 Galium odoratum group18 Circaea x intermedia 1.1 Oxalis acetosella group19 Convallaria majalis 1.1 Oxalis acetosella group20 Corydalis cava 1.1 Mercurialis perennis group21 Crepis paludosa 2.1 Galium odoratum group22 Dactylorhiza fuchsii 2.1 Mercurialis perennis group23 Epipactis helleborine 1.1 Oxalis acetosella group24 Equisetum hyemale 1.1 Galium odoratum group25 Equisetum pratense 1.1 Galium odoratum group26 Equisetum sylvaticum 1.1 Galium odoratum group27 Equisetum telmateia 1.1 Mercurialis perennis group28 Festuca altissima 1.1 Galium odoratum group29 Gagea spathacea 1.1 Galium odoratum group30 Galium odoratum 1.1 Galium odoratum group31 Geum rivale 2.1 Mercurialis perennis group32 Gymnocarpium dryopteris 1.1 Oxalis acetosella group33 Hordelymus europaeus 1.1 Mercurialis perennis group34 Hypericum pulchrum 2.1 Oxalis acetosella group35 Ilex aquifolium 1.1 Oxalis acetosella group36 Impatiens noli-tangere 1.1 Galium odoratum group37 Lamium galeobdolon 1.1 Galium odoratum group38 Listera ovata 1.1 Mercurialis perennis group39 Luzula pilosa 1.1 Galium odoratum group40 Luzula sylvatica subsp. sylvatica 1.1 Mercurialis perennis group41 Lysimachia nemorum 1.1 Galium odoratum group42 Maianthemum bifolium 1.1 Oxalis acetosella group43 Melica uniflora 1.1 Galium odoratum group44 Mercurialis perennis 1.1 Mercurialis perennis group45 Milium effusum 1.1 Galium odoratum group46 Neottia nidus-avis 1.1 Galium odoratum group47 Orchis mascula 2.1 Mercurialis perennis group48 Oreopteris limbosperma 1.1 Oxalis acetosella group49 Oxalis acetosella 1.1 Oxalis acetosella group50 Paris quadrifolia 1.1 Galium odoratum group51 Phegopteris connectilis 1.1 Oxalis acetosella group52 Phyteuma spicatum 1.1 Mercurialis perennis group53 Platanthera chlorantha 1.1 Galium odoratum group54 Potentilla sterilis 1.2 Mercurialis perennis group55 Primula elatior 1.1 Galium odoratum group56 Pulmonaria obscura 1.1 Galium odoratum group57 Ranunculus auricomus agg. 2.1 Galium odoratum group58 Ranunculus lanuginosus 1.1 Mercurialis perennis group59 Rumex sanguineus 1.1 Galium odoratum group60 Sanicula europaea 1.1 Galium odoratum group61 Scrophularia nodosa 2.1 Oxalis acetosella group62 Scutellaria galericulata 2.1 Oxalis acetosella group63 Stachys sylvatica 1.1 Galium odoratum group64 Ulmus laevis 1.1 Oxalis acetosella group65 Veronica montana 1.1 Galium odoratum group66 Viola reichenbachiana 1.1 Galium odoratum group67 Viola riviniana 1.1 Galium odoratum group

234 M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239

1.1) was highest in the clusters of ancient woodland species(G. odoratum, M. perennis, and O. acetosella groups). In contrast,indifferent species (R. ficaria, D. flexuosa, and G. hederacea groups)and recent woodland species (A. capillaris group) grew predominantlyin forests, as well as in open areas (category 2.1). The proportion ofvascular plants largely restricted to closed forests (category 1.1)was lowest in the A. capillaris group of recent woodland species.In this group, however, there was a prominence of speciespreferring forest edges and clearings (category 1.2).

3.2. Compilation of an implementable ancient woodland indicatorplant list

The list of ancient woodland indicator plants (Table 3; AppendixTable A1) comprised 67 species. Most of them (85%) belonged tothe forest species category 1.1 (largely restricted to closed forests),13% were part of the category 1.2 (preferring forest edges andclearings), and 2% belonged to the category 2.1 (occurring inforests, as well as in open land).

Fig. 6 displays the numbers of ancient woodland indicator plantspecies (forest species categories 1.1 and 1.2) present in the gridquadrants for northwest Germany. The highest numbers of indica-tor species were found in Schleswig–Holstein, where the easternhill country as a young moraine landscape emerges. The smalland fragmented ancient woodlands, which are predominant in thisregion, are characterised by nutrient-rich soils (Niemann, 1809;Hase, 1997). Such nutrient-rich sites support the highest diversityof ancient woodland plant species (e.g., Wulf, 2004).

4. Discussion

4.1. Woodland plant species groups and their ecological characteristics

Of the seven woodland plant species groups identified for ourstudy area, we found three main groups of ancient woodland

species; the G. odoratum, M. perennis, and O. acetosella groups.The species composition of these groups is substantially in linewith the results of other studies conducted in temperate Western

Fig. 5. Linkage to forest habitats within the seven plant groups identified by k-means clustering. 1.1 – largely restricted to closed forests, 1.2 – preferring forest edges andclearings, 2.1 – occurring in forests, as well as in open areas.

Fig. 6. Number of ancient woodland indicator plant species per topographic map quadrant in northwest Germany considering only forest species of categories 1.1 (largelyrestricted to closed forests) and 1.2 (preferring forest edges and clearings).

M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239 235

and Central Europe (e. g., Wulf, 1997, 2004; Hermy et al., 1999;Verheyen et al., 2003). Of particular note is the G. odoratum group,where 71% of the plant species we recorded are also listed byHermy et al. (1999) as ancient forest species. In the M. perennisgroup, this is true for 62% of the species and in the O. acetosellagroup for only 41% of the species. On the one hand, the relativelylow concordance of the latter group may possibly be explainedby the low number of studies examining ancient woodland plantson acidic soils (see Heinken, 1998; Matuszkiewicz et al., 2013).Thus, Hermy et al. (1999), whose study was dependent on the

availability of local ancient woodland plant species lists, could onlydraw conclusions for a locally limited species pool. On the otherhand, considering z-values, the affinity of acidophytic plant speciescharacterising this group to deciduous ancient woodland sitesseems to be generally less pronounced. This may be related tothe ability of many of these species to grow in coniferous or mixedforests or, alternatively, to persist in extensively managed openhabitats (forest affinity category 2.1; Schmidt et al., 2011) suchas heathlands composed of dwarf shrubs or matt-grass swards(Wulf, 2004; Ellenberg and Leuschner, 2010).

236 M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239

Our analysis further revealed that, even among the plantsstrictly bound to forests, there is a group of species strictly linkedto recent woodlands. However, this group, the A. capillaris group, isrelatively small and includes only common and very commonspecies. Furthermore, we assume that a larger number of plantslinked to recent woodlands is contained in the disregarded groupof plants that may occur in forests, but preferably grow in openareas (forest affinity category 2.2; Schmidt et al., 2011). Similarresults were found for the Prignitz region, which adjoins our studyarea to the east (Wulf, 2004).

However, most of the forest plants included in our analysisbelong to one of the three groups of species that are more or lessindifferent to woodland continuity (i.e. the R. ficaria, G. hederacea,and D. flexuosa groups). These groups contain very common andubiquitous forest plant species without any linkage to ancientwoodlands. Furthermore, there are many very rare speciesincluded, whose possible (local or regional) linkage to ancientwoodlands could not be statistically verified on the supra-regionalscale of our study.

The results of the k-means cluster analysis allowed for areasoned interpretation of the ecological conditions with which thespecies groups are linked based on mean Ellenberg indicator values(EIV). The gradient from ancient to recent woodland species groupscorresponded particularly positively to an increase in the EIV forlight. We conclude that the light demand of plant species plays acrucial role for their linkage to ancient or recent woodland sites(see also Petersen, 1994; Howard and Lee, 2003). Shade-tolerantforest species mostly belonged to one of the three groups identifiedas ancient woodland species. This is also reflected by the highproportion of plant species restricted to closed forests. In contrast,both the species group linked to recent woodland sites and thethree groups of species more or less indifferent to woodlandcontinuity are characterised by a high light demand and theycontain only a few species that are restricted to closed forests.Our results are very much in line with those of Hermy et al. (1999)and Wulf (2004), who concluded that vascular plants characteristicfor ancient deciduous woodlands are more shade-tolerant thanother forest plant species. However, British ancient woodlandsare often characterised by more light-demanding woodland plantsdue to the long history of coppicing (Kirby, 1990; Rose, 1999).

The results of the EIV for soil reaction, which are correlated tothe second axis of the PCA, require a more complex ecologicalinterpretation. The species with the highest demand for basesupply occur in the M. perennis group, followed by the G. odoratumgroup. Both groups are highly correlated to deciduous woodlandon ancient woodland sites (as indicated by high z-values). Incontrast, the O. acetosella group (i.e. the third group of ancientwoodland species) and the three groups with more or less indifferentwoodland continuity all show a wide range of EIV for soil reaction.Finally, the A. capillaris group of recent woodland species ischaracterised by both a low need for base saturation and a narrowrange of EIV for soil reaction. From the latter, we conclude thatafforestation and recent natural forest development in our studyarea occurred particularly on sites with acidic soils (cf. Kremser,1990; Hase, 1997). Hermy et al. (1999) found that ancientwoodland plants, when compared to other forest plants, mostlyoccur in woodlands with both intermediate pH values and nitrogenavailability, and are in most cases lacking on both dry and wetsites. We could confirm their results in respect to base saturation,but not in respect to nitrogen supply (as quantified by the EIV)where we found only a very weak relationship to the groups ofancient woodland species. Although for soil moisture, we foundalmost no relationship, there was a distinct set of indicator speciesfor wet soil conditions, such as Chrysosplenium oppositifolium,Crepis paludosa or Equisetum telmateia, which were closely linkedto ancient woodland sites, and Convallaria majalis, Orchis mascula

and Viola riviniana, a set of ancient woodland plants known to befound on drier sites (Ellenberg et al., 2001).

4.2. A new approach in determining ancient woodland indicator plantlists and their potential application

In Europe, lists of ancient woodland indicator plants have upuntil now almost exclusively been compiled by expert knowledgeor were based on vegetation relevés in combination with local orregional species lists (Hermy et al., 1999; Verheyen et al., 2003;Perrin and Daly, 2010). Such lists have also been compiled forsome smaller areas of northwest Germany (Wulf, 2004). As anexpectable consequence of this approach, regional differences inthe linkage of plant species to ancient woodlands were stronglyemphasised (Wulf, 1997; Kühn, 2000). Therefore, there has beena need to develop a more robust and standardised approach tocompile lists of ancient woodland indicator plants (Glaves et al.,2009), which are implementable on a supra-regional scale. Wehave applied a new approach by using repeatable statisticalmethods and large comprehensive species distribution data setsin combination with archive records on ancient woodlands forthe determination of ancient woodland indicator plant lists. Indoing so, we have used additional information on the ecology ofthe plant species. Firstly, we reduced the list to the set of plantspecies most closely linked to woodland habitats based oninformation available for all of Germany and perhaps beyond(Schmidt et al., 2011), and, secondly, we considered the preferenceof the forest plant species for deciduous, coniferous, or mixedwoodland on ancient woodland sites. However, when consideringthe z-values, it becomes obvious that there are no absolute limitsfor the classification of a plant species as an ancient woodland indi-cator. There is a rather continuous transition from species closelylinked to ancient woodland sites to species with a linkage to recentwoodland sites. Hence, we developed the procedure described inchapter 2.3.2 and compiled an implementable ancient woodlandindicator plant list. This resulted in a list of 67 significant ancientwoodland indicator plants with supra-regional validity and appli-cability. We believe that our approach is easily applicable to otherlarge areas in Europe, because high-resolution data on historicaland recent land cover and the distribution of vascular plant speciesare becoming increasingly available in more and more countries(see Table 6 in Culmsee et al., 2014). However, it has to be consid-ered that there are not any vascular plants that grow exclusively onancient woodland sites (Rose, 1999; Glaves et al., 2009). In order toidentify an ancient woodland site with high accuracy, one has todetect multiple ancient woodland indicator plants (Rose, 1999;Kühn, 2000; Schmidt et al., 2009). With regard to the necessarynumber of ancient woodland indicator plant species, the valuesin the literature range from 2 (Kühn, 2000) to 27 (Honnay et al.,1998). Honnay et al. (1998) stressed that the indicative value ofancient woodland plant species is scale dependent. Furtherresearch is needed in this area.

Widely applicable ancient woodland indicator plant listsmay be a useful tool for nature conservation practice, where thepotential applications are:

(a) Identification of ancient woodlands in areas where historicalmaps are lacking

There are regions where historical maps are completely lackingor where it is difficult to obtain them from archives (Rose, 1999;Crawford, 2009). For instance, the German-wide ancient woodlandinventory provided by Glaser and Hauke (2004) does not cover theGerman federal state of Hamburg as a part of northwest Germany.For this area, the occurrence of indicator species can give evidenceof ancient woodland sites. The same is true for the region of

M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239 237

southern Denmark adjacent to our study area. Ancient woodlandindicators may also be useful when ancient woodland inventoriesare, due to small map resolution, less accurate for smaller woodlandpatches or woodland fringes (Wulf, 2004; Goldberg et al., 2007;Oheimb et al., 2007). In addition, ancient woodland indicators areimportant for the identification of ‘‘ancient hedges’’, which areremnants of, or were once adjacent to, original forests (Pollardet al., 1974; Stone and Williamson, 2013). Such old linear landscapestructures can serve as propagule sources for the spread of ancientwoodland species into adjacent recent woodlands (Corbit et al.,1999; Liira and Paal, 2013; Stone and Williamson, 2013).

(b) Identification of biodiversity hotspots of ancient woodlandindicator plants

Not every ancient woodland site shows a high species richnessin ancient woodland indicator plants. In fact, there may be a largevariation in alpha diversity due to land-use history, silviculturaltreatment, tree species composition, and nutrient supply(Dupouey et al., 2002; Härdtle et al., 2003; Mölder et al., 2014b).Following the ‘‘hotspot strategy’’ of Meyer et al. (2009), woodlandpatches with a high diversity of ancient woodland indicator plantsshould be identified and protected or managed for natureconservation (Schmiedel et al., 2013).

(c) Ancient woodland indicator plants as indicators for ancientsemi-natural woodlands

Since ancient woodland indicator plants allow for the detectionof woodland sites with long habitat continuity, under certainconditions, they can also serve as indicators for ancient semi-naturalwoodland and natural diversity (Rose, 1999; Nordén andAppelqvist, 2001; D’Amato et al., 2009). Habitat continuity, whenaccompanied by structural continuity (e. g., the occurrence ofancient trees), allows for conclusions to be drawn on the wholecommunity of woodland species including bryophytes, lichens,fungi, or beetles (Ferris and Humphrey, 1999; Grove, 2002;Kriebitzsch et al., 2013; Mölder et al., 2014a). In this context,ancient woodland indicators can be part of a mapping procedurefor the identification of ancient semi-natural woodland; especially,when historical maps are difficult to obtain (Rose, 1999; Nordénand Appelqvist, 2001; Crawford, 2009; Goldberg et al., 2007).

5. Conclusions

In our opinion, for nature conservation practice, there is a greatneed for ancient woodland indicator plant lists with supra-regionalapplicability. In this study, we introduce a new methodicalapproach that allows the compilation of such lists by using readilyavailable resources of plant species monitoring programs and landcover data. Using the area of northwest Germany as a modelregion, we presented an ecologically grounded ancient woodlandindicator plant list. In northwest Germany, where only 26% of thewoodlands are ancient, the proportion of deciduous ancientwoodland sites amounts to merely 7% and these woodlands arehighly fragmented. In such scarcely wooded, agriculturallydominated landscapes, the value of forest islands for natureconservation depends on historical ecological continuity (Thomaset al., 1997). Under such circumstances, deciduous ancient woodlandsites can be hotspots of forest plant biodiversity. Furthermore, theycan act as propagule sources for the spread of ancient woodlandspecies into adjacent recent woodlands. However, in Europe, thetime for such a spread takes up to 350–800 years (Falinski, 1986;Peterken, 1977; Rackham, 2003).

In order to promote the typical plant diversity of ancientsemi-natural woodlands, effective conservation management

should strongly support the preservation of ancient deciduouswoodlands and inhibit their conversion to coniferous or mixedstands. The connection of recent and ancient woodlands by habitatcorridors should be strengthened (Hermy et al., 1999; De Frenneet al., 2011; Kriebitzsch et al., 2013; Verstraeten et al., 2013;Leuschner et al., 2014). In fragmented landscapes with intenseagriculture, no-spray buffer zones of at least 5 m should beadopted to protect the majority of woodland species from theimpacts of agrochemicals applied to adjacent land (Gove et al.,2007). Furthermore, the forest management of deciduous ancientwoodland sites with a high typical woodland plant diversity hasto be carefully conducted to avoid soil damage (Worrell andHampson, 1997; Godefroid and Koedam, 2004). These actions mustbe taken in stands within protected areas (Thomas et al., 1997;Schmiedel et al., 2013), but should also be promoted beyond, sinceexisting protected area networks usually cover only part of theecologically valuable ancient woodlands in which forest floordiversity is particularly difficult to restore (Thompson et al.,2003; Thomas et al., 1997; De Frenne et al., 2011).

Acknowledgments

This study was made possible by innumerable volunteers andprofessionals who reported plant species occurrences in the statesof Lower Saxony, Bremen and Schleswig-Holstein. We thankAnnemarie Schacherer (Lower Saxony Water Management, CoastalDefence and Nature Conservation Agency) and Katrin Romahn(AG Geobotanik in Schleswig-Holstein und Hamburg e. V.) forproviding floristic data. We gratefully acknowledge the funding ofthe projects ‘‘Identification of indicator species groups of grasslandand forest habitats for biodiversity monitoring and evaluation’’(Grant Number DBU 26752) and ‘‘Identification and protection offorest stands of special importance for biodiversity conservation’’(Grant Number DBU 29677) by the German Federal Foundationfor the Environment (DBU). We thank Ruth Gilbert and Bob Larkinfor proofreading. We are also indebted to two anonymousreviewers for suggestions that have greatly improved the paper.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.foreco.2014.06.043.

References

AG Geobotanik, 2013. Gemeinsame Datenbank der AG Geobotanik in Schleswig-Holstein und Hamburg e. V. und des Landesamtes für Landwirtschaft, Umweltund ländliche Räume des Landes Schleswig-Holstein. Kiel and Flintbek.

Arnold, V., 2011. Celtic fields und andere urgeschichtliche Ackersysteme inhistorisch alten Waldstandorten Schleswig-Holsteins aus Laserscan-Daten.Archäol. Korr.bl. 41, 439–455.

Brock, G., Pihur, V., Datta, S., Datta, S., 2008. ClValid: an R package for clustervalidation. J. Stat. Softw. 25, 1–22.

Corbit, M., Marks, P.L., Gardescu, S., 1999. Hedgerows as habitat corridors for forestherbs in central New York, USA. J. Ecol. 87, 220–232. http://dx.doi.org/10.1046/j.1365-2745.1999.00339.x.

Coppins, A.M., Coppins, B.J., 2002. Indices of Ecological Continuity for WoodlandEpiphytic Lichen Habitats in the British Isles. British Lichen Society, Edinburgh.

Crawford, C., 2009. Ancient woodland indicator plants in Scotland. Scott. For. 63, 6–19.

Culmsee, H., Schmidt, M., Schmiedel, I., Schacherer, A., Meyer, P., Leuschner, C.,2014. Predicting the distribution of forest habitat types using indicator speciesto facilitate systematic conservation planning. Ecol. Indic. 37, 131–144. http://dx.doi.org/10.1016/j.ecolind.2013.10.010 (Part A).

D’Amato, A.W., Orwig, D.A., Foster, D.R., 2009. Understory vegetation in old-growthand second-growth Tsuga canadensis forests in western Massachusetts. For.Ecol. Manage. 257, 1043–1052. http://dx.doi.org/10.1016/j.foreco.2008.11.003.

De Frenne, P., Baeten, L., Graae, B.J., Brunet, J., Wulf, M., Orczewska, A., Kolb, A.,Jansen, I., Jamoneau, A., Jacquemyn, H., Hermy, M., Diekmann, M., De Schrijver,A., De Sanctis, M., Decocq, G., Cousins, S.A.O., Verheyen, K., 2011. Interregional

http://dx.doi.org/10.1016/j.foreco.2014.06.043http://dx.doi.org/10.1016/j.foreco.2014.06.043http://refhub.elsevier.com/S0378-1127(14)00422-8/h0010http://refhub.elsevier.com/S0378-1127(14)00422-8/h0010http://refhub.elsevier.com/S0378-1127(14)00422-8/h0010http://refhub.elsevier.com/S0378-1127(14)00422-8/h0015http://refhub.elsevier.com/S0378-1127(14)00422-8/h0015http://dx.doi.org/10.1046/j.1365-2745.1999.00339.xhttp://dx.doi.org/10.1046/j.1365-2745.1999.00339.xhttp://refhub.elsevier.com/S0378-1127(14)00422-8/h0025http://refhub.elsevier.com/S0378-1127(14)00422-8/h0025http://refhub.elsevier.com/S0378-1127(14)00422-8/h0030http://refhub.elsevier.com/S0378-1127(14)00422-8/h0030http://dx.doi.org/10.1016/j.ecolind.2013.10.010http://dx.doi.org/10.1016/j.ecolind.2013.10.010http://dx.doi.org/10.1016/j.foreco.2008.11.003

238 M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239

variation in the floristic recovery of post-agricultural forests. J. Ecol. 99, 600–609. http://dx.doi.org/10.1111/j.1365-2745.2010.01768.x.

Diekmann, M., Duprè, C., Kolb, A., Metzing, D., 2008. Forest vascular plants asindicators of plant species richness: a data analysis of a flora atlas fromnorthwestern Germany. Plant Biosyst. 142, 584–593. http://dx.doi.org/10.1080/11263500802410934.

Dubberke-Spandlowski, M., 2011. Alte Waldstandorte in den LandesforstenSchleswig-Holsteins. Gutachten im Auftrag des NiedersächsischenForstplanungsamtes (NFP). Gesellschaft für Forstplanung GbR, Salzgitter.

Dupouey, J.L., Dambrine, E., Laffite, J.D., Moares, C., 2002. Irreversible impact of pastland use on forest soils and biodiversity. Ecology 83, 2978–2984. http://dx.doi.org/10.1890/0012-9658(2002) 083[2978:IIOPLU]2.0.CO;2.

Dzwonko, Z., Gawroński, S., 1994. The role of woodland fragments, soil types, anddominant species in secondary succession on the western Carpathian foothills.Vegetatio 111, 149–160. http://dx.doi.org/10.1007/BF00040334.

Ellenberg, H., Leuschner, C., 2010. Vegetation Mitteleuropas mit den Alpen inökologischer, dynamischer und historischer Sicht, sixth ed. Ulmer, Stuttgart.

Ellenberg, H., Weber, H.E., Düll, R., Wirth, V., Werner, W., 2001. Zeigerwerte vonPflanzen in Mitteleuropa. Scr. Geobot. 18, 1–264.

Fahrmeir, L., Kneib, T., Lang, S., 2009. Regression: Modelle, Methoden undAnwendungen, second ed. Springer, Berlin and Heidelberg.

Falinski, J.B., 1986. Vegetation Dynamics in Temperate Lowland Primeval Forests.Dr. W. Junk Publishers, Dordrecht, Boston, Lancaster.

Ferris, R., Humphrey, J.W., 1999. A review of potential biodiversity indicators forapplication in British forests. Forestry 72, 313–328. http://dx.doi.org/10.1093/forestry/72.4.313.

Garve, E., Schacherer, A., Bruns, E., Feder, J., Täuber, T., 2007. Verbreitungsatlas derFarn- und Blütenpflanzen in Niedersachsen und Bremen. Nat.schutzLandsch.pfl. Niedersachs. 43, 1–507.

Glaser, F.F., Hauke, U., 2004. Historisch alte Waldstandorte und Hutewälder inDeutschland. Angew. Landsch.ökol. 61, 1–193.

Glaves, P., Rotherham, I.D., Wright, B., Handley, C., Birbeck, J., 2009. A survey of thecoverage, use and application of ancient woodland indicator lists in the UK.Hallam Environmental Consultants Ltd., Biodiversity and Landscape HistoryResearch Institute/Geography, Tourism and Environment Change Research Unit,Sheffield Hallam University, Sheffield.

Godefroid, S., Koedam, N., 2004. Interspecific variation in soil compaction sensitivityamong forest floor species. Biol. Conserv. 119, 207–217. http://dx.doi.org/10.1016/j.biocon.2003.11.009.

Goldberg, E., Kirby, K., Hall, J., Latham, J., 2007. The ancient woodland concept as apractical conservation tool in Great Britain. J. Nat. Conserv. 15, 109–119. http://dx.doi.org/10.1016/j.jnc.2007.04.001.

Gove, B., Power, S.A., Buckley, G.P., Ghazoul, J., 2007. Effects of herbicide spray drift andfertilizer overspread on selected species of woodland ground flora: comparisonbetween short-term and long-term impact assessments and field surveys. J. Appl.Ecol. 44, 374–384. http://dx.doi.org/10.1111/j.1365-2664.2007.01261.x.

Grove, S.J., 2002. Saproxylic insect ecology and the sustainable management offorests. Annu. Rev. Ecol. Syst. 33, 1–23. http://dx.doi.org/10.1146/annurev.ecolsys.33.010802.150507.

Härdtle, W., von Oheimb, G., Westphal, C., 2003. The effects of light and soilconditions on the species richness of the ground vegetation of deciduous forestsin northern Germany (Schleswig-Holstein). For. Ecol. Manage. 182, 327–338.http://dx.doi.org/10.1016/S0378-1127(03)00091-4.

Hase, W., 1997. Wald- und Forstchronologie Schleswig-Holsteins seit derNacheiszeit. Struve’s Buchdruckerei und Verlag, Eutin.

Heinken, T., 1998. Zum Einfluss des Alters von Waldstandorten auf die Vegetation inbodensauren Laubwäldern des niedersächsischen Tieflandes. Arch. Nat.schutzLandsch.forsch. 37, 201–232.

Hermy, M., Honnay, O., Firbank, L., Grashof-Bokdam, C., Lawesson, J.E., 1999. Anecological comparison between ancient and other forest plant species ofEurope, and the implications for forest conservation. Biol. Conserv. 91, 9–22.http://dx.doi.org/10.1016/S0006-3207(99)00045-2.

Hermy, M., Verheyen, K., 2007. Legacies of the past in the present-day forestbiodiversity: a review of past land-use effects on forest plant speciescomposition and diversity. In: Nakashizuka, T. (Ed.), Sustainability andDiversity of Forest Ecosystems. Springer, Japan, pp. 361–371.

Honnay, O., Degroote, B., Hermy, M., 1998. Ancient forest plant species in westernBelgium: a species list and possible ecological mechanisms. Belg. J. Bot. 130,139–154.

Howard, L.F., Lee, T.D., 2003. Temporal patterns of vascular plant diversity insoutheastern New Hampshire forests. For. Ecol. Manage. 185, 5–20. http://dx.doi.org/10.1016/S0378-1127(03)00243-3.

Ito, S., Nakayama, R., Buckley, G., 2004. Effects of previous land-use on plant speciesdiversity in semi-natural and plantation forests in a warm-temperate region insoutheastern Kyushu, Japan. For. Ecol. Manage. 196, 213–225. http://dx.doi.org/10.1016/j.foreco.2004.02.050.

Kelemen, K., Kriván, A., Standovár, T., 2014. Effects of land-use history and currentmanagement on ancient woodland herbs in Western Hungary. J. Veg. Sci. 25,172–183. http://dx.doi.org/10.1111/jvs.12046.

Kimberley, A., Blackburn, G.A., Whyatt, J.D., Kirby, K., Smart, S.M., 2013. Identifyingthe trait syndromes of conservation indicator species: how distinct are Britishancient woodland indicator plants from other woodland species? Appl. Veg. Sci.16, 667–675. http://dx.doi.org/10.1111/avsc.12047.

Kirby, K.J., 1990. Changes in the ground flora of a broadleaved wood within a clearfell, group fells and a coppiced block. Forestry 63, 241–249. http://dx.doi.org/10.1093/forestry/63.3.241.

Kremser, W., 1990. Niedersächsische Forstgeschichte: Eine integrierteKulturgeschichte des nordwestdeutschen Forstwesens. SelbstverlagHeimatbund Rotenburg (Wümme), Rotenburg (Wümme).

Kriebitzsch, W.-U., Bültmann, H., Oheimb, G.V., Schmidt, M., Thiel, H., Ewald, J., 2013.Forest-specific diversity of vascular plants, bryophytes, and lichens. In: Kraus, D.,Krumm, F. (Eds.), Integrative Approaches as an Opportunity for the Conservationof Forest Biodiversity. European Forest Institute, Joensuu, pp. 158–169.

Kühn, I., 2000. Ökologisch-numerische Untersuchungen an Wäldern in derWestfälischen Bucht – Ein Beitrag zur Biodiversitäts- und Altwald-Forschung.Arch. Naturwiss. Diss. 12, 1–192.

Leuschner, C., Wulf, M., Bäuchler, P., Hertel, D., 2014. Forest continuity as a keydeterminant of soil carbon and nutrient storage in beech forests on sandy soilsin Northern Germany. Ecosystems 17, 497–511. http://dx.doi.org/10.1007/s10021-013-9738-0.

Liira, J., Paal, T., 2013. Do forest-dwelling plant species disperse along landscapecorridors? Plant Ecol. 214, 455–470. http://dx.doi.org/10.1007/s11258-013-0182-1.

Matuszkiewicz, J.M., Kowalska, A., Kozłowska, A., Roo-Zielińska, E., Solon, J., 2013.Differences in plant-species composition, richness and community structure inancient and post-agricultural pine forests in central Poland. For. Ecol. Manage.310, 567–576. http://dx.doi.org/10.1016/j.foreco.2013.08.060.

Meyer, P., Schmidt, M., Spellmann, H., 2009. Die ‘‘Hotspots-Strategie’’ – Wald-Naturschutzkonzept auf landschaftsökologischer Grundlage. AFZ/Wald 64,822–824.

Mölder, A., Gürlich, S., Engel, F., 2014a. Die Verbreitung von gefährdeten Holzbewohnenden Käfern in Schleswig-Holstein unter dem Einfluss vonForstgeschichte und Besitzstruktur. Forstarchiv 85, 84–101. http://dx.doi.org/10.4432/0300-4112-85-84.

Mölder, A., Streit, M., Schmidt, W., 2014b. When beech strikes back: How strictnature conservation reduces herb-layer diversity and productivity in CentralEuropean deciduous forests. For. Ecol. Manage. 319, 51–61. http://dx.doi.org/10.1016/j.foreco.2014.01.049.

Moning, C., Müller, J., 2009. Critical forest age thresholds for the diversity of lichens,molluscs and birds in beech (Fagus sylvatica L.) dominated forests. Ecol. Indic. 9,922–932. http://dx.doi.org/10.1016/j.ecolind.2008.11.002.

Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000.Biodiversity hotspots for conservation priorities. Nature 403, 853–858. http://dx.doi.org/10.1038/35002501.

Nascimbene, J., Thor, G., Nimis, P.L., 2013. Effects of forest management onepiphytic lichens in temperate deciduous forests of Europe – a review. For. Ecol.Manage. 298, 27–38. http://dx.doi.org/10.1016/j.foreco.2013.03.008.

Niemann, A.C.H., 1809. Forststatistik der dänischen Staten. Johann FriedrichHammerich, Altona.

Nordén, B., Appelqvist, T., 2001. Conceptual problems of ecological continuity andits bioindicators. Biodivers. Conserv. 10, 779–791. http://dx.doi.org/10.1023/A:1016675103935.

Oheimb, G. von, Schmidt, M., Kriebitzsch, W.-U., 2007. Waldflächenentwicklung imöstlichen Schleswig-Holstein in den letzten 250 Jahren und ihre Bedeutung fürseltene Gefäßpflanzen. Tuexenia 27, 363–380.

Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B.,Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H., 2012. Vegan:Community Ecology Package.

Perrin, P.M., Daly, O.H., 2010. A provisional inventory of ancient and long-established woodland in Ireland. Irish Wildlife Manuals 46, 1–65.

Peterken, G.F., 1974. A method for assessing woodland flora for conservation usingindicator species. Biol. Conserv. 6, 239–245. http://dx.doi.org/10.1016/0006-3207(74)90001-9.

Peterken, G.F., 1977. Habitat conservation priorities in British and Europeanwoodlands. Biol. Conserv. 11, 223–236. http://dx.doi.org/10.1016/0006-3207(77)90006-4.

Petersen, P.M., 1994. Flora, vegetation, and soil in broadleaved ancient and plantedwoodland, and scrub on Røsnæs, Denmark. Nord. J. Bot. 14, 693–709. http://dx.doi.org/10.1111/j.1756-1051.1994.tb01086.x.

Pollard, E., Hooper, M.D., Moore, N.W., 1974. Hedges. Collins, London.QGIS Development Team, 2014. QGIS Geographic Information System. Open Source

Geospatial Foundation Project. .R Development Core Team, 2013. R: A language and environment for statistical

computing. R Foundation for Statistical Computing, Vienna, .Raabe, E.-W., 1987. Atlas der Flora Schleswig-Holsteins und Hamburgs. Wachholtz-

Verlag, Neumünster.Rackham, O., 2003. Ancient Woodland: Its History, Vegetation and Uses in England,

new ed. Castlepoint Press, Dalbeattie.Rose, F., 1999. Indicators of ancient woodland – the use of vascular plants in

evaluating ancient woods for nature conservation. Brit. Wildlife 10, 241–251.Schmidt, M., Kriebitzsch, W.-U., Ewald, J., 2011. Waldartenlisten der Farn- und

Blütenpflanzen, Moose und Flechten Deutschlands. BfN-Skripten 299, 1–111.Schmidt, M., Meyer, P., Paar, U., Evers, J., 2009. Bedeutung der Habitatkontinuität für

die Artenzusammensetzung und -vielfalt der Waldvegetation. Forstarchiv 80,195–202.

Schmiedel, I., Schmidt, M., Schacherer, A., Culmsee, H., 2013. Die Effektivität vonSchutzgebieten für die Erhaltung seltener und gefährdeter Gefäßpflanzenarten– Eine Untersuchung im niedersächsischen Tiefland. Nat.schutz Landsch.plan.45, 45–52.

Singleton, R., Gardescu, S., Marks, P.L., Geber, M.A., 2001. Forest herb colonization ofpostagricultural forests in central New York State, USA. J. Ecol. 89, 325–338.http://dx.doi.org/10.1046/j.1365-2745.2001.00554.x.

http://dx.doi.org/10.1111/j.1365-2745.2010.01768.xhttp://dx.doi.org/10.1080/11263500802410934http://dx.doi.org/10.1080/11263500802410934http://dx.doi.org/10.1890/0012-9658(2002)083[2978:IIOPLU]2.0.CO;2http://dx.doi.org/10.1890/0012-9658(2002)083[2978:IIOPLU]2.0.CO;2http://dx.doi.org/10.1007/BF00040334http://refhub.elsevier.com/S0378-1127(14)00422-8/h0070http://refhub.elsevier.com/S0378-1127(14)00422-8/h0070http://refhub.elsevier.com/S0378-1127(14)00422-8/h0075http://refhub.elsevier.com/S0378-1127(14)00422-8/h0075http://refhub.elsevier.com/S0378-1127(14)00422-8/h0080http://refhub.elsevier.com/S0378-1127(14)00422-8/h0080http://refhub.elsevier.com/S0378-1127(14)00422-8/h0085http://refhub.elsevier.com/S0378-1127(14)00422-8/h0085http://dx.doi.org/10.1093/forestry/72.4.313http://dx.doi.org/10.1093/forestry/72.4.313http://refhub.elsevier.com/S0378-1127(14)00422-8/h0095http://refhub.elsevier.com/S0378-1127(14)00422-8/h0095http://refhub.elsevier.com/S0378-1127(14)00422-8/h0095http://refhub.elsevier.com/S0378-1127(14)00422-8/h0100http://refhub.elsevier.com/S0378-1127(14)00422-8/h0100http://dx.doi.org/10.1016/j.biocon.2003.11.009http://dx.doi.org/10.1016/j.biocon.2003.11.009http://dx.doi.org/10.1016/j.jnc.2007.04.001http://dx.doi.org/10.1016/j.jnc.2007.04.001http://dx.doi.org/10.1111/j.1365-2664.2007.01261.xhttp://dx.doi.org/10.1146/annurev.ecolsys.33.010802.150507http://dx.doi.org/10.1146/annurev.ecolsys.33.010802.150507http://dx.doi.org/10.1016/S0378-1127(03)00091-4http://refhub.elsevier.com/S0378-1127(14)00422-8/h0135http://refhub.elsevier.com/S0378-1127(14)00422-8/h0135http://refhub.elsevier.com/S0378-1127(14)00422-8/h0140http://refhub.elsevier.com/S0378-1127(14)00422-8/h0140http://refhub.elsevier.com/S0378-1127(14)00422-8/h0140http://dx.doi.org/10.1016/S0006-3207(99)00045-2http://refhub.elsevier.com/S0378-1127(14)00422-8/h0155http://refhub.elsevier.com/S0378-1127(14)00422-8/h0155http://refhub.elsevier.com/S0378-1127(14)00422-8/h0155http://dx.doi.org/10.1016/S0378-1127(03)00243-3http://dx.doi.org/10.1016/S0378-1127(03)00243-3http://dx.doi.org/10.1016/j.foreco.2004.02.050http://dx.doi.org/10.1016/j.foreco.2004.02.050http://dx.doi.org/10.1111/jvs.12046http://dx.doi.org/10.1111/avsc.12047http://dx.doi.org/10.1093/forestry/63.3.241http://dx.doi.org/10.1093/forestry/63.3.241http://refhub.elsevier.com/S0378-1127(14)00422-8/h0190http://refhub.elsevier.com/S0378-1127(14)00422-8/h0190http://refhub.elsevier.com/S0378-1127(14)00422-8/h0190http://refhub.elsevier.com/S0378-1127(14)00422-8/h0190http://refhub.elsevier.com/S0378-1127(14)00422-8/h0195http://refhub.elsevier.com/S0378-1127(14)00422-8/h0195http://refhub.elsevier.com/S0378-1127(14)00422-8/h0195http://dx.doi.org/10.1007/s10021-013-9738-0http://dx.doi.org/10.1007/s10021-013-9738-0http://dx.doi.org/10.1007/s11258-013-0182-1http://dx.doi.org/10.1007/s11258-013-0182-1http://dx.doi.org/10.1016/j.foreco.2013.08.060http://refhub.elsevier.com/S0378-1127(14)00422-8/h0215http://refhub.elsevier.com/S0378-1127(14)00422-8/h0215http://refhub.elsevier.com/S0378-1127(14)00422-8/h0215http://dx.doi.org/10.4432/0300-4112-85-84http://dx.doi.org/10.4432/0300-4112-85-84http://dx.doi.org/10.1016/j.foreco.2014.01.049http://dx.doi.org/10.1016/j.foreco.2014.01.049http://dx.doi.org/10.1016/j.ecolind.2008.11.002http://dx.doi.org/10.1038/35002501http://dx.doi.org/10.1038/35002501http://dx.doi.org/10.1016/j.foreco.2013.03.008http://refhub.elsevier.com/S0378-1127(14)00422-8/h0245http://refhub.elsevier.com/S0378-1127(14)00422-8/h0245http://dx.doi.org/10.1023/A:1016675103935http://dx.doi.org/10.1023/A:1016675103935http://refhub.elsevier.com/S0378-1127(14)00422-8/h0255http://refhub.elsevier.com/S0378-1127(14)00422-8/h0255http://refhub.elsevier.com/S0378-1127(14)00422-8/h0255http://refhub.elsevier.com/S0378-1127(14)00422-8/h0265http://refhub.elsevier.com/S0378-1127(14)00422-8/h0265http://dx.doi.org/10.1016/0006-3207(74)90001-9http://dx.doi.org/10.1016/0006-3207(74)90001-9http://dx.doi.org/10.1016/0006-3207(77)90006-4http://dx.doi.org/10.1016/0006-3207(77)90006-4http://dx.doi.org/10.1111/j.1756-1051.1994.tb01086.xhttp://dx.doi.org/10.1111/j.1756-1051.1994.tb01086.xhttp://refhub.elsevier.com/S0378-1127(14)00422-8/h0285http://qgis.osgeo.orghttp://www.R-project.orghttp://refhub.elsevier.com/S0378-1127(14)00422-8/h0300http://refhub.elsevier.com/S0378-1127(14)00422-8/h0300http://refhub.elsevier.com/S0378-1127(14)00422-8/h0305http://refhub.elsevier.com/S0378-1127(14)00422-8/h0305http://refhub.elsevier.com/S0378-1127(14)00422-8/h0310http://refhub.elsevier.com/S0378-1127(14)00422-8/h0310http://refhub.elsevier.com/S0378-1127(14)00422-8/h0315http://refhub.elsevier.com/S0378-1127(14)00422-8/h0315http://refhub.elsevier.com/S0378-1127(14)00422-8/h0320http://refhub.elsevier.com/S0378-1127(14)00422-8/h0320http://refhub.elsevier.com/S0378-1127(14)00422-8/h0320http://refhub.elsevier.com/S0378-1127(14)00422-8/h0325http://refhub.elsevier.com/S0378-1127(14)00422-8/h0325http://refhub.elsevier.com/S0378-1127(14)00422-8/h0325http://refhub.elsevier.com/S0378-1127(14)00422-8/h0325http://dx.doi.org/10.1046/j.1365-2745.2001.00554.x

M. Schmidt et al. / Forest Ecology and Management 330 (2014) 228–239 239

Spencer, J.W., Kirby, K.J., 1992. An inventory of ancient woodland for England andWales. Biol. Conserv. 62, 77–93. http://dx.doi.org/10.1016/0006-3207(92)90929-H.

Stone, A., Williamson, T., 2013. ‘‘Pseudo-Ancient Woodland’’ and the AncientWoodland Inventory. Landscapes 14, 141–154. http://dx.doi.org/10.1179/1466203513Z.00000000016.

Svenning, J.-C., Baktoft, K.H., Balslev, H., 2008. Land-use history affects understoreyplant species distributions in a large temperate-forest complex, Denmark.Plant Ecol. 201, 221–234. http://dx.doi.org/10.1007/s11258-008-9557-0.

Szabó, P., 2009. Open woodland in Europe in the Mesolithic and in the Middle Ages:can there be a connection? For. Ecol. Manage. 257, 2327–2330. http://dx.doi.org/10.1016/j.foreco.2009.03.035.

Thomas, R.C., Kirby, K.J., Reid, C.M., 1997. The conservation of a fragmentedecosystem within a cultural landscape – the case of ancient woodland inEngland. Biol. Conserv. 82, 243–252. http://dx.doi.org/10.1016/S0006-3207(97)00039-6.

Thompson, R.N., Humphrey, J.W., Harmer, R., Ferris, R., 2003. Restoration of NativeWoodland on Ancient Woodland Sites – Forestry Commission Practice Guide.Forestry Commission, Edinburgh.

Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics with S, fourth ed.Springer, New York.

Verheyen, K., Honnay, O., Motzkin, G., Hermy, M., Foster, D.R., 2003. Response offorest plant species to land-use change: a life-history trait-based approach. J.Ecol. 91, 563–577. http://dx.doi.org/10.1046/j.1365-2745.2003.00789.x.

Verstraeten, G., Baeten, L., De Frenne, P., Vanhellemont, M., Thomaes, A., Boonen, W.,Muys, B., Verheyen, K., 2013. Understorey vegetation shifts following theconversion of temperate deciduous forest to spruce plantation. For. Ecol.Manage. 289, 363–370. http://dx.doi.org/10.1016/j.foreco.2012.10.049.

Wisskirchen, R., Haeupler, H., 1998. Standardliste der Farn- und BlütenpflanzenDeutschlands. Ulmer, Stuttgart.

Worrell, R., Hampson, A., 1997. The influence of some forest operations on thesustainable management of forest soils – a review. Forestry 70, 61–85. http://dx.doi.org/10.1093/forestry/70.1.61.

Wulf, M., 1997. Plant species as indicators of ancient woodland in northwesternGermany. J. Veg. Sci. 8, 635–642. http://dx.doi.org/10.2307/3237367.

Wulf, M., 2003. Preference of plant species for woodlands with differing habitatcontinuities. Flora 198, 444–460. http://dx.doi.org/10.1078/0367-2530-00118.

Wulf, M., 2004. Auswirkungen des Landschaftswandels auf die Verbreitungsmustervon Waldpflanzen. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart.

http://dx.doi.org/10.1016/0006-3207(92)90929-Hhttp://dx.doi.org/10.1016/0006-3207(92)90929-Hhttp://dx.doi.org/10.1179/1466203513Z.00000000016http://dx.doi.org/10.1179/1466203513Z.00000000016http://dx.doi.org/10.1007/s11258-008-9557-0http://dx.doi.org/10.1007/s11258-008-9557-0http://dx.doi.org/10.1016/j.foreco.2009.03.035http://dx.doi.org/10.1016/j.foreco.2009.03.035http://dx.doi.org/10.1016/S0006-3207(97)00039-6http://dx.doi.org/10.1016/S0006-3207(97)00039-6http://refhub.elsevier.com/S0378-1127(14)00422-8/h0360http://refhub.elsevier.com/S0378-1127(14)00422-8/h0360http://refhub.elsevier.com/S0378-1127(14)00422-8/h0360http://refhub.elsevier.com/S0378-1127(14)00422-8/h0365http://refhub.elsevier.com/S0378-1127(14)00422-8/h0365http://dx.doi.org/10.1046/j.1365-2745.2003.00789.xhttp://dx.doi.org/10.1016/j.foreco.2012.10.049http://refhub.elsevier.com/S0378-1127(14)00422-8/h0380http://refhub.elsevier.com/S0378-1127(14)00422-8/h0380http://dx.doi.org/10.1093/forestry/70.1.61http://dx.doi.org/10.1093/forestry/70.1.61http://dx.doi.org/10.2307/3237367http://dx.doi.org/10.1078/0367-2530-00118http://refhub.elsevier.com/S0378-1127(14)00422-8/h0400http://refhub.elsevier.com/S0378-1127(14)00422-8/h0400

Determining ancient woodland indicator plants for practical use: A new approach developed in northwest Germany1 Introduction2 Materials and Methods2.1 Study area2.2 Data sets2.3 Statistical analysis2.3.1 Identification of supra-regional ancient woodland plants2.3.2 Compilation of an ancient woodland indicator plant list for practical use

3 Results3.1 Identification of supra-regional ancient woodland plants3.2 Compilation of an implementable ancient woodland indicator plant list

4 Discussion4.1 Woodland plant species groups and their ecological characteristics4.2 A new approach in determining ancient woodland indicator plant lists and their potential application

5 ConclusionsAcknowledgmentsAppendix A Supplementary materialReferences