Trade-off between light availability and soil fertility determine refugial conditions for the relict light-demanding species in lowland forests

Acta Oecologica 85 (2017) 1e8

Contents lists available at ScienceDirect

Acta Oecologica journal homepage: www.elsevier.com/locate/actoec

Trade-off between light availability and soil fertility determine refugial conditions for the relict light-demanding species in lowland forests
ski a, *, Jo
zef Krzysztof Kurowski a, Edyta Kiedrzyn
ska b, c Marcin Kiedrzyn
d
z, ul Banacha 12/16, 90-237 Ło
d
z, Poland Department of Geobotany and Plant Ecology, Faculty of Biology and Environmental Protection, University of Ło
d
z, Poland European Regional Centre for Ecohydrology of the Polish Academy of Sciences, ul Tylna 3, 90-364 Ło c
d
z, ul. Banacha 12/16, 90-237 Ło
d
z, Poland Department of Applied Ecology, Faculty of Biology and Environmental Protection, University of Ło a

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 27 February 2017 Received in revised form 5 September 2017 Accepted 5 September 2017

Identifying potential refugial habitats in the face of rapid environmental change is a challenge faced by scientists and nature conservation managers. Relict populations and refugial habitats are the model objects in those studies. Based on the example of Actaea europaea from Central Poland, we analyse the habitat factors influencing relict populations of continental, light-demanding species in lowland forests and examine which habitats of studied species corresponding most closely to ancient vegetation. Our results indicate that the current refugial habitats of Actaea europaea include not only communities which are very similar to ancient open forest but also forests with a closed canopy. Although the populations are influenced by nitrogen and light availability, the co-occurrence of these two factors in forest communities is limited by dense canopy formation by hornbeam and beech trees on fertile soils and in more humid conditions. Our findings indicate that the future survival of relict, light-demanding communities in lowland forests requires low-intensity disturbances to be performed in tree-stands, according to techniques, which imitate traditional forests management. © 2017 Elsevier Masson SAS. All rights reserved.

Keywords: Habitat suitability Habitat factors interaction Relict populations Continental species

1. Introduction The identification of places or habitats suitable for biodiversity protection during rapid environmental change is a challenge faced by scientists and nature conservation managers (Ashcroft, 2010; Keppel et al., 2012). One such area of scientific importance whose significance has greatly increased over recent decades is that of refugial sites: habitats which could harbour populations in the region under unsuitable environmental conditions (Keppel et al.,
ski et al., 2017). In the light of current climate 2015; Kiedrzyn change, relicts and communities associated with them are unique natural laboratories and model systems for observation (Hampe and Jump, 2011; Woolbright et al., 2014). It is well known that topographical complexity generates more possibilities for the formation of numerous in situ refugial habitats under climate change (Keppel et al., 2012), while plain areas tend to encourage shifts in species range to follow suitable conditions

* Corresponding author.
ski). E-mail address: [email protected] (M. Kiedrzyn http://dx.doi.org/10.1016/j.actao.2017.09.004 1146-609X/© 2017 Elsevier Masson SAS. All rights reserved.

(Moritz and Agudo, 2013). While a species exposed to climate change may survive in a rugged landscape with significantly varied abiotic conditions, biotic interactions mostly drive the formation of refugial habitats for populations in plain landscapes, especially in lowland areas (Hampe and Jump, 2011). Examples of habitats for relict populations of mountain plants in the European lowland
ski areas have been given previously (Grzyl et al., 2014; Kiedrzyn
ska et al., 2016). The aim of the present paper is et al., 2015; Zielin to analyse the conditions which drive the existence of continental light-demanding plant in the area of the closed-canopy forests e zonal vegetation in the Central European lowlands. One of the patterns described for the glacial/interglacial cycles is species migration along the oceanic-continental gradient (Stewart et al. 2010). ‘Oceanic’ adaptation implies a more humid, less seasonably variable climate, and ‘continental’ adaptation a drier climate with greater seasonal variation (Ellenberg, 1988). In Central Europe, such long-term vegetation dynamics along the oceaniccontinental gradient are mostly in a longitudinal direction (Stewart et al. 2010). During glacial periods, continental species from the Eastern Europe have been widely dispersed to the west, according to the distribution of the steppe-like vegetation or hemi-

2


ski et al. / Acta Oecologica 85 (2017) 1e8 M. Kiedrzyn

boreal forests (van Andel and Tzedakis, 1996; Birks and Willis, 2008). During interglacial periods, including the Holocene, continental light-demanding species have been in decline in Central Europe, when the transition from late-glacial cold steppe and open woodlands to the mixed and closed forests occurred (Puhe and Ulrich, 2001; Chytrý et al., 2010). As a result, some continental species already exist only as isolated populations in the region, and their relict status has been revealed recently by phylogeographic studies (Stewart et al. 2010; Kajtoch et al. 2016 and cited literature). Moreover, future patterns associated with climate change in Central Europe provide an eastward shift of conditions suitable for oceanic plants (Theurillat and Guisan, 2001; Bakkenes et al., 2002; Skov and Svenning, 2004; Thuiller et al., 2005). Hence, this projected future trends follow the long-term west-east shifts of species associated with the previous glacial and interglacial climate oscillations in the region (Stewart et al., 2010). Of the continental plants, strictly steppic species are currently found in Central Europe mostly in extrazonal xeric grasslands, and many aspects concerning their ecology, phylogeography and conservation have been studied (Kajtoch et al., 2016). Less is known of the light-demanding continental species which exist mainly in forest or shrub communities. They are regarded today as climate relicts of late-glacial open woods (Ellenberg, 1988), and maybe more importantly, as biotic relicts which have lost their sites as a result of the Holocene expansion of mesophilic trees (RalskaJasiewiczowa et al., 2003). Subsequent invasions by elm, maple, lime, hornbeam and beech since the mid-Holocene reduced light availability below the canopy, resulting in a dramatic decline in the numbers of light-demanding species from forest communities (Chytrý et al., 2010). Some continental species, adapted to existence in nutrient-poor soils (e.g. Pulsatilla patens, Arenaria graminifolia), could benefit from the relatively higher light availability in that habitats, structured by coniferous or mixed stands. In fertile habitats, the shading pressure from broadleaved trees is stronger, and the balance between soil fertility and light availability drives the existence of light-demanding plants. They forced then to occur in sub-optimal habitats and play the “habitat trade-off game”: they are forced to continually choose between light availability and other habitat resources needed for their effective growth and reproduction. Such species include for example Adenophora liliifolia and Lathyrus pisiformis: indicators of thermophilous forest hotspots
ski and which are currently disappearing from the region (Kiedrzyn Jakubowska-Gabara, 2014). Our study examines the pattern described above, using the example of another subcontinental plant, Actaea europaea (Schipcz.) J. Compton (Ranununculaceae), a survivor of early-Holocene open forest (Rolecek, 2007) which grows on rather fertile soils. Actaea europaea is the only taxon in Europe from the section Cimicifuga (Compton and Hedderson, 1997), and is closely related to A. cimicifuga from Siberia, despite being separated by a distance of 4000 km. This disjunction is probably a result of the Pleistocene extinction of the ancestor of A. europaea (Compton et al., 1998), implying that its current range could be recognized as a relict distribution on a broad Euro-Asiatic scale. In addition, the European-scale relict character is indicated by its current diffuse distribution (Fig. 1a). Hence, considering the historic and biogeographic aspects, the species is recognized as a relict in Central Europe. The predicted distribution of suitable conditions for A. europaea suggest that within the next few decades, the suitable climatic conditions of the species will shift north-east with climate change (Skov and Svenning, 2004). Unfortunately, with the longterm restriction of suitable habitats for continental species observed since the mid-Holocene and the projected unfavourable perspectives in Central Europe, the future of A. europaea in the region is rather pessimistic.

Populations of A. europaea on the western border of the temperate continental bioclimate in plain landscape of Central Poland were chosen for this study (Fig. 1a), as such border localities in lowlands are under particular threat from projected climate change. In the study area, A. europaea have been noted in various
ski and forest habitats on mesic and eutrophic soils (Kiedrzyn Andrzejewski, 2012). Surviving populations can be found in open oak forests, but also in hornbeam and beech stands, which are rather closed by their nature. To determine which factors play a crucial role in the actual well-being of populations, our study goes on to examine various aspects of the current habitats of A. europaea, including soil parameters and community compositions. It could be hypothesized that none of the current habitats of A. europaea in Central Poland exactly recreate the ancient conditions of the species. However, our analysis also examines which current habitat of A. europaea corresponds most closely to the ancient habitats suitable for a light-demanding continental species. Hence, the forest communities with A. europaea from Central Poland have been compared to southern Ural forest communities, which are regarded as being analogous to early- and mid-Holocene Central European forest communities (Chytrý et al., 2010). Modern analogues of ancient Central European forests occur in the region which are currently outside the geographical ranges of beech and hornbeam and where the current continental climate resembles the early Holocene climate of Central Europe (Davis et al., 2003). The analysis includes open woodlands of coniferous and smallleaved deciduous trees, which are considered as analogues to early-Holocene vegetation, as well as broad-leaved deciduous forests and closed-canopy forests, which are analogues to midHolocene vegetation in Central Europe (Willis and van Andel, 2004; Birks and Willis, 2008). Our analysis, therefore, allows the refugial habitats of A. europaea in Central Poland to be compared with habitats equivalent to the temporal sequence of the postglacial habitats. The main aims of the study were: 1) to detect the basic habitat parameters influencing the current populations of subcontinental species A. europaea in Central Poland; 2) to analyse possible opposed interactions between environmental parameters in the current lowland refugial habitats of A. europaea; 3) to indicate the similarity of the present plant communities with A. europaea with potential late-glacial vegetation by analogy with the southern Ural forests.

2. Materials and methods 2.1. Study area The study area includes the central part of the Polish section of the western border of the geographical range of A. europaea in Europe (Fig. 1a). The average annual temperature in the region ranges from 7.4 to 7.8
C and the average annual precipitation from 565 to 625 mm (data from period 1982e2012, Climate-Data.org, 2016). The main geological units on the surface are Pleistocene clays, sands and gravels from the Odranian and the Wartanian glaciations (corresponding to the Saale glaciation; 300 to 130 Kya) (Marks et al., 2006). However, Mesozoic rocks (limestones, sandstones) have a visible share in the southern and the south-eastern parts of the region (Marks et al., 2006). The populations of Actaea europaea are thought to be disappearing from the region, and the species was included to the
ski and Andrzejewski, 2012), with only regional red book (Kiedrzyn


ski et al. / Acta Oecologica 85 (2017) 1e8 M. Kiedrzyn

3

Fig. 1. The range of Actaea europaea in Europe (a) according to: Meusel et al. (1965); Jalas and Souminen (1989); Zaja˛ c and Zaja˛ c (2001); www.nature.portal.cz, changed and supplemented; range of temperate continental bioclimate according to Rivas-Martínez et al., (2010); (b) occurrence of A. europaea in Central Poland; populations found and studied

w, CZ e Czartoria, MP e Miejskie Pola, BK e Bukowa Go
ra, GK e Go
ra Krzemyk, WS e Wola Swidzi
ska, KO e Korzecko. in 2013: TE e Teofilo n

ten populations confirmed in the study area since 2005, and a few extinct populations being described (Fig. 1b). The occurrence of preserved A. europaea localities is signifi
ski cantly associated with that of limestone outcrops (Kiedrzyn et al., 2016); these are located between 146 and 320 m a.s.l. and coincide with the location of hotspots of thermophilous forest flora
ski and Jakubowska-Gabara, 2014). A. europaea grows in (Kiedrzyn subcontinental dry-mesic oak forests (Potentillo albae-Quercetum Libb. 1933), central European thermophilous beech forests (Cephalanthero-Fagenion Tx. 1955; Fagus sylvatica-Cruciata glabra community) and thermophilous variants of subcontinental oakhornbeam forests (Tilio-Carpinetum Tracz. 1962; subass. melittietosum).

2.2. Population size and generative potential The basic population parameters was determined in 2013, when only seven populations were found. The total size of the population was determined in September, when the species is fully generative developing in the region, by counting the total number of individuals in the population. Each individual was regarded as a separate plant including individuals with a few or numerous shoots. The number of individuals was counted in two categories: vegetative (non-flowering) and generative (flowering) plants. During the field study, it was found that within the studied localities of A. europaea the tree-layer had not been managed for over 10 years. Hence, the current structure and conditions were typical of the forest type. However, any previous disturbances (if occurred recently) should be visible in the structure and the composition of plant communities, which were analysed in the study.

2.3. Vegetation survey in localities The basic vegetation survey of current A. europaea habitats was conducted in 2013 in each studied locality. One vegetation plot
) was conducted for each locality, and performed in such as a (releve way as to cover the greatest possible number of individuals of A. europaea and to be representative of the vegetation type. The
s was 400 m2. However in most cases, the area area of the releve occupied by a population of A. europaea in a given locality was
s). In each locality, larger than the area of vegetation plots (releve the populations occupied a single habitat type; therefore, it is enough to use one representative plot for vegetation type analysis. The floristic composition and structure of plant communities were
s using the phytosociological method (Braunsurveyed in releve Blanquet, 1964). The nomenclature used for the vascular plants was in accordance with the accepted names from the Plant List (www.theplantlist.org; version 1.1, September 2013).
s were analysed with Juice 7.0.105 softThe vegetation releve ware (Tichý and Holt, 2006). The unweighted mean habitat indi
s was calculated based on Ellenberg indicator cator values for releve values for Light and Moisture (Ellenberg et al., 1992). The calculated Light and Moisture indexes were then used in the multidimensional analysis (PCA) of habitats.

2.4. Soil sample and analysis The soil samples were taken from all studied localities of A. europaea in Central Poland from the top level of the profile up to a depth of 20 cm (mostly 5e15 cm). As the top 5 cm of the forest floor was composed of litter which had not uniformly accumulated,

4


ski et al. / Acta Oecologica 85 (2017) 1e8 M. Kiedrzyn

the soil samples were taken after removing the upper layer of the soil. The samples were taken from three points within each locality and were then uniformed. After preparation of the soil samples, exchangeable (pHKCl) acidity was determined potentiometrically. The percentage content of organic matter was determined using the loss-on-ignition method, the percentage content of total nitrogen using Kiejdahl's method, available K2O and P2O2 according to Egner-Riehm, and Mg content using the atomic absorption method. All analyses were performed in an accredited laboratory according to Polish National Procedures (presented in Supplement 1). As differences in organic matter content were found between samples, modifications were made to the procedures for determining the available potassium, phosphorus and magnesium content in the organic soil samples taken from the rendzinas soils on limestones, where the organic matter content was greater than 10%. In order to compare the results from different kinds of soil, the percentage of mean values (presented in Supplement 1) for the particular types of soil in Poland were calculated.


s from the four types southern Ural area, which included 78 releve of forest, was obtained from Chytrý et al. (2010). The analysis includes 35 species, which represent the core floristic compositions of the analysed forests. The percentage of
s with analysed species in a particular forest type, i.e. the releve frequency, was used as a measure of the species association to the particular forest type. Consequently, the analysis includes species whose frequency was not lower than 10% in at least one forest type (Supplement 2). The similarity between forest types was assessed by hierarchical cluster analysis, with the Euclidian distance taken as a dissimilarity measure and the Ward's method as a clustering method (Statistica 10 package, Statsoft 2011). In addition, an ordination of forest types from Central Poland and the southern Urals was also performed. The ‘cases’ were forest types and ‘variables’ were the frequency of the species. As the data set had a gradient length of 0.7 SD units, unconstrained and linear PCA analysis was run together with the log-transformation and centring of response data. Ordination and  graphic plots were constructed using Canoco 5 software (Smilauer and Leps, 2014).

2.5. Analysis of habitat factors 3. Results The correlation between the population parameters of A. europaea and the habitat factors was evaluated using the Spearman's rank correlation coefficient. The Statistica 10 package (Statsoft, 2011) was used for the calculations. The factors most significant for A. europaea populations were then used for multidimensional ordination (Principal Component Analysis PCA) of localities according to habitat conditions. The ‘cases’ were localities and ‘variables’ were the habitat factors. In addition to significant soil parameters, the indicator values (Light
s were also used. index, Moisture index) calculated from the releve Unconstrained and linear PCA analysis was used by centring and standardizing data. The analysis was performed and graphic plots  constructed using Canoco 5 software (Smilauer and Leps, 2014). 2.6. Similarity of current communities to ancient vegetation Actaea europaea is known to be a survivor of early-Holocene open forests, which were forced to cede their territory to the expansion of deciduous broad-leaved trees (Role cek, 2007). As early- and mid-Holocene Central European vegetation is regarded as being analogous to modern southern Ural forest communities (Chytrý et al., 2010), the study compares communities with A. europaea from Central Poland to southern Ural forests. It is reasonable to view the ancient Holocene communities as analogous to the extant communities in Poland (Chytrý et al., 2010) as although A. europaea is endemic for Central Europe and does not occur in the Ural region, the forests of the Ural region contain most of the species known from the Central European forests. The analysis allowed similarities to be found between the current refugial habitats of A. europaea and the potential ancient vegetation. Because not all species occur in both regions, only the species shared between Central Poland and southern Ural were used in the analysis.
s with A. europaea from three types of The analysis used 17 releve forest in Central Poland. The data set from Central Poland included
s performed by the authors and 10 releve
s with seven releve A. europaea taken from published materials (Supplement 2). These additional materials were taken to increase the number of cases for analysis due to the limited number of sites which were available for study in Central Poland at the time. It should be noted that the beech and hornbeam forests were analysed together, according to preliminary PCA ordination, which indicated that these forest types demonstrated close floristical similarity. The data from the

3.1. Most important factors influencing populations In the study area, the populations in beech and oak forests tended to have higher numbers of individuals (Table 1). The percentage of generative plants is the highest in oak forests (40e60%) (Table 1). In general, the percentage of generative individuals was positively related to the number of individuals in the population (rank Spearman correlation coefficients ¼ 0.67) but the value was not statistically significant. A significant positive relationship was found between the total number of individuals in a population and total nitrogen content (0.82) (Table 2). In addition, a significant positive relationship (0.79) was found between light availability and the share of flowering plants in the population. Some factors have clear relations to population parameters; however, the results were statistically not significant (Table 2). The number of individuals in a population was positively related with organic matter content and available forms of phosphorus, potassium and magnesium, as well as light availability. Available magnesium content was found to be positively related to both population parameters, while moisture index was negatively and insignificantly related to population parameters. Although these relationships are not significant, probably due to the low number of cases, they are highlighted here because they could have an important influence on the studied populations. 3.2. Differences in habitat conditions The significant habitat resources are shared between different habitats (Fig. 2). The PCA 1 axis (accounting for 50.3% of the variation) could be interpreted as a gradient running from subcontinental oak forest with higher light level and magnesium availability to the relatively moist and shaded hornbeam forest (Fig. 2, Table 3). The second PCA axis (30.5%) indicated a fertility gradient running from eutrophic beech to mesotrophic oak forests on sands (Fig. 2, Table 3). However, hornbeam forests are dispersed along this axis. 3.3. Similarity of the current communities to the ancient forest In the hierarchical clustering model, the oak forests with Actaea europaea from Central Poland were found in the same cluster as open pine-larch, birch-poplar and oak forests from the southern


ski et al. / Acta Oecologica 85 (2017) 1e8 M. Kiedrzyn

5

Table 1 General characteristics of studied localities of Actaea europaea on the western border of its range in Central Poland. Bold values indicate the share of generative individuals above 40%. Locality names according to Fig. 1. Forest type

Dry-mesic oak forest Potentillo albae-Quercetum

Thermophilous beech forest Cephalanthero-Fagenion

Thermophilous hornbeam forest Tilio-Carpinetum melittietosum

Parent material of soil

Sand, Gravel

Limestone

Lime-stone, Sand

Sand-stone

Lime-stone

Locality name Tree-layer cover [%] Shrub-layer cover [%] Number of individuals Vegetative Generative (% in population) Total

MP 50 50

KO 60 30

BK 60 60

GK 60 40

TE 70 40

CZ 70 50

WS 60 70

140 211 (60.1) 351

31 22 (41.5) 53

310 23 (6.9) 333

258 106 (29.1) 364

17 0 (0%) 17

4 0 (0%) 4

61 9 (12.8) 70

Table 2 The rank Spearman correlation coefficients between population and habitat parameters in localities of Actaea europaea. Bold values are significant at p < 0.05. Soil parameters: Org% e percentage of organic matter; Ntot% e percentage of total nitrogen content; P/K/Mg/%mean e percentage of available phosphorus/potassium/magnesium/content
s according to Ellenberg indicator values for plant species. according to mean values for regional soils. Ecological indicators: Light index, Moisture index e counted for the releve Population parameters

pH

Org %

Ntot %

P2O5 %mean

K2O %mean

MgO %mean

Light index

Moisture index

Total number of individuals p Share of generative individuals p

0.18 0.701 0.32 0.478

0.71 0.071 0.18 0.699

0.82 0.023 0.34 0.452

0.46 0.294 0.16 0.728

0.56 0.192 0.02 0.969

0.50 0.253 0.45 0.310

0.57 0.180 0.79 0.033

0.32 0.482 0.45 0.310

Fig. 2. PCA ordination of localities of Actaea europaea in Central Poland by habitat conditions. Forest types: black square e dry-mesic oak forest, triangle e thermophilous beech forest, circle e thermophilous variant of oak-hornbeam forest. The numbers in brackets correspond to the population size. Abbreviations of soil parameters are given according to Table 2 and abbreviations for locality names according to Fig. 1.

Urals (Fig. 3a). The second main cluster included more closed maple-lime-elm forests from the Urals and beech with hornbeam forests from Poland (Fig. 3a). The results of the PCA revealed groups based on geographic region and ecological differentiation. The first PCA axis, accounting for 45.0% of the variation, indicated that the present-day oak forest from Central Poland was similar to the pine-larch, birch-poplar and oak forest types from the southern Urals. Some indicator species for these groups of forest were found to be light-demanding plants: Sanquisorba officinalis, Rubus saxatilis, Galim boreale, Calamagrostis arundinacea, Brachypodium pinnatum or Veronica chamaedrys (Fig. 3b; Supplement 3). In turn, forests with beech and hornbeam from Central Poland are more similar to the maple-lime-elm forest type from the Urals, which are analogous to later stages of central European vegetation in the Holocene. Species related to the above habitats include: Galium odoratum, Tilia cordata, Campanula trachelium, Acer platanoides or Aegopodium podagraria. The second PCA axis, accounting for 29.9% of the total variation, clearly divided the Central Poland and southern Ural communities, and could be interpreted as being based on the geographic differentiation (continental gradient) of communities (Fig. 3b). The responses of species on the first two PCA axes are presented in Supplement 3. 4. Discussion

Table 3 Responses on the first two PCA axes in ordination of localities of Actaea europaea in central Poland by chosen habitat parameters. Bold values indicate the habitat parameters, which are the most correlated to PCA axes (positively or negatively and r coefficient above 0.8). Abbreviations of soil parameters are given according to Table 2; r e regression coefficient. PCA1 (50.3%)

PCA2 (30.5%)

Habitat condition

r

Habitat condition

r

Moisture index pH Ntot% Light index MgO%mean

0.86 0.15 0.26 ¡0.87 ¡0.96

Ntot% pH Moisture index MgO%mean Light index

0.85 0.84 0.11 0.03 0.26

The spread of broad-leaved trees in Central Europe during the mid-Holocene caused the formation of a dense canopy in forests and a resulting decline of the light-demanding continental species from the herb layer (Chytrý et al., 2010). Actaea europaea is considered to be one such species, and it probably lost a large area of suitable habitats at that time. As shown by our results, relatively high light availability and soil fertility are the most significant factors influencing A. europaea in current refugial habitats. In the study area, the co-occurrence of these two factors in forest communities is limited by dense canopy formation by hornbeam and beech trees on fertile soils. Nevertheless, they still have an influence on the total number of individuals in populations of A. europaea in Central Poland and the proportions of flowering plants.

6


ski et al. / Acta Oecologica 85 (2017) 1e8 M. Kiedrzyn

Fig. 3. Similarity between forests with Actaea europaea from Central Poland and forests from the southern Urals; (a) cluster analysis using Euclidean Distance and Ward's method; (b) PCA ordination diagram: solid dots e Central Poland forests, empty dots e Ural forests. Analysis performed using values of frequency in the communities of 35 species mutual for both regions. The species names and their responses on the PCA axes are shown in Supplement 3.

Rather large populations of 50e350 individuals with the highest ratios of generative plants (40e60%) were found in dry-mesic oak forests on mesotrophic sandy soils, which are associated with higher light availability. Our results indicate that the closest floristic match for these Central European oak forest community is represented by the species-rich open pine-larch and birch-poplar and oak forests from the southern Urals. Although this is consistent with previous suggestions by Chytrý et al. (2010), our study is the first to offer support based on numerical analysis. Hemiboreal pinelarch forests predominate in the summit areas and on the eastern (Siberian; dryer and more continental) side of the Southern Ural range. Birch-aspen forests often occur at previously disturbed sites, but they can also form natural forests in some places, especially on the Siberian side of the mountains. Oak forests are found as zonal vegetation in drier areas at the south-eastern fringes of the mountain range and can also occur at drier or formerly disturbed sites within the western (European) zone of the Ural Mountains (Popov, 1980; Martynenko et al., 2005). These three types of open forest communities include numerous light-demanding species with relicts of steppe communities and are regarded as analogous to the early-Holocene forest of Central Europe (Chytrý et al., 2010). The species composition of dry-mesic oak forests currently found in Central Europe on mesotrophic soils was mostly maintained by stable light availability over the last few centuries resulting in pressure from humans and large herbivores (Vera,
dl et al., 2010). The ref2000; Ralska-Jasiewiczowa et al., 2003; He ugial role of oak forests is supported by the relict species occurrence (Rolecek, 2007) and a relatively large species pool, which
dlo et al., 2007). includes numerous light-demanding plants (Sa The following features appear to be conducive for the refugial role played by dry-mesic oak forests, allowing them to harbour the relict populations of such continental species as A. europaea. Although the herb layer in oak forests is exposed to relatively more light and more radiant heat, they actually have intermediate microclimatic and edaphic conditions (Ellenberg, 1988; Babi
c et al., 2010; Szymura and Szymura, 2011). In Poland, they occur in welldrained sand and gravel deposits with deep groundwater levels
ska, 1973; Dego
rski, 1990). Our (Matuszkiewicz and Roo-Zielin

present findings suggest that the dry-mesic oak forests are the driest habitat in the study area in which A. europaea occurs. This dryness could be a limiting factor for large herbs such as A. europaea. However, in this type of habitat, base-rich rocks (e.g. limestones in the study area), clays or loam occur often under the loose deposits and impair water drainage, resulting in higher soil moisture levels, especially in the first part of the season
ska, 1973; Dego
rski, 1990; Chytrý, (Matuszkiewicz and Roo-Zielin 1997; Role cek, 2007); this phenomenon fosters conditions suitable for large perennials such as A. europaea. Large populations (330e360 individuals) were found to occur also in beech forest on limestone rendzinas, with high organic matter content, and high levels of nitrogen and phosphorus; however, these tend not to have a high share of flowering individuals (6e29%). In turn, smaller populations of four to 70 individuals occur in hornbeam forests, on rather eutrophic soils and in a relatively moist habitat, and these generally lack flowering plants, mainly due to the low availability of light. As indicated by our results, an inverse relationship exists between moisture and light availability in the current habitats of A. europaea in Central Poland; in more humid habitats (especially in hornbeam forest), a denser canopy has a higher chance of formation. The forests with beech and hornbeam stands from the study area have quite similar species composition, with the underground dominated by Central European herbs, typical of closed forests. However, the patches examined in the present study represent thermophilous variants of beech and hornbeam forests, and apart from A. europaea, some other light-demanding species also occur
ski and Andrzejewski, 2012). Our results also within them (Kiedrzyn indicate that the species composition of both forests is similar to that of the maple-lime-elm forests from the Southern Ural, which serve as analogues to the later stage of vegetation evolution during the Holocene in the Central Europe (Chytrý et al., 2010). They reflect existing zonal vegetation present in habitats with higher pH and more nutrient-rich soils as well as with faster decomposition of litter (Chytrý et al., 2010). This structure and biochemical pattern in communities, together with the lower light availability, have


ski et al. / Acta Oecologica 85 (2017) 1e8 M. Kiedrzyn

resulted in a lower abundance of light-demanding and steppic herbaceous plants, as in the case of studied beech and hornbeam forests from Central Poland. Hence, these type of forests harbour much lower numbers of light-demanding continental and steppic plants than dry-mesic oak forests (our results and compare with
dlo et al., 2007). Sa The long-term existence of relict populations of sub-continental plants such as A. europaea in more eutrophic and humid habitats is possible in regions with sharper topography, i.e. on the steep slopes of valleys, ravines or hills (Compton and Hedderson, 1997; Chytrý
dlo, 1997; Jutrzenka-Trzebiatowski and Dziedzic, 1998). and Sa These sites can offer an adequate level of light due to disturbances resulting from mass wasting processes and tree saltation: Such a “cyclic forest-free” refugial habitat is found in the Czech Republic, on the western border of the range of A. europaea. In such habitats, re-succession of the trees follows disturbances, with the process
dlo, 1997). repeated over time (Chytrý and Sa In a flat landscape, human activity can caused disturbances in the tree-stands and helped some light-demanding species to survive. In Central Poland, the presence of A. europaea in beech and hornbeam forests is attributed to human activities with two key aspects. The first aspect is the combination of a reduction of the area of the forest complex and tree-stand management, resulting in the creation of an intensive forest-edge effect (Bomanowska and
ski, 2011). Actaea europaea especially benefits from this Kiedrzyn effect in beech complexes, with only small patches of being preserved in the study area. A second factor is the establishment and extensive use of small quarries in forest complexes in the vicinity of A. europaea localities, which could have continued for centuries
ski, 2009). This is a consequence of the fact that in the (Kiedrzyn region, the species is associated with outcrops of Mesozoic rocks, mostly limestones. In conclusion, beech and hornbeam forests could offer sub-optimal conditions for A. europaea in Central Poland on calcareous soils and if light conditions are maintained by creating disturbances in tree stands. The disappearance of A. europaea after the formation of dense
ski tree-stands is well documented in Central Poland (Kiedrzyn and Andrzejewski, 2012). In this regard, even subcontinental oak-forests are currently not stable habitats. After the cessation of traditional forest management, hornbeam, linden and hazel were able to invade over the period of a few decades (Jakubowska-Gabara, 1996). This process is relatively fast in the more fertile and humid variants of thermophilous forests, which are paradoxically more suitable for A. europaea in the region. Similar habitat preferences have been described for relict populations of Adenophora liliifolia, also light-demanding continental species which occur in Central European forests (Parusov
a et al., 2016). Habitats which could act as long-term ‘safe havens’ against climate or environmental change are not believed to be as common in flat landscapes as in more rugged areas (Keppel et al., 2012). Hence, the threat of species extinction is higher in lowland areas, and especially in the anthropogenic landscape, where the fragmentation of habitats greatly hampers species migration (Young et al., 1996). In the case of A. europaea, long-term traditional management of open woodlands could help the species survival for centuries. However, the intensive forestry which has taken place in the study area since the 19th Century, caused the fragmentation of even those cultural habitats. Too much intensive forest management, associated with the sudden opening of the habitats, and dense plantations can damage individual A. europaea plants. As a result, almost all populations of the species in the region are currently isolated. With the restriction of species dispersion in mind (Compton et al., 1998), it can be assumed that the future survival of the relict populations of

7

A. europaea in Central Poland, would be more likely to occur in situ, if at all possible. 5. Conclusions None of the present-day habitats of Actaea europaea in Central Poland offer optimal conditions for its survival, even oak forests, whose habitat is the most similar to that of ancient open forests. None of the studied habitats represents a self-sufficient ‘save haven’ for A. europaea and the future existence of the studied relict populations requires the active creation of habitat structure. This structure and refugial function of the habitats is derived from interactions between light availability, soil fertility and moisture. Low-intensity disturbances which imitate traditional forest management need to be created in tree-stands to ensure the optimal combination of these conditions. This would allow for the existence of key elements of light-demanding forest communities on mesic and eutrophic habitats, which harbour an important and disappearing part of regional biodiversity. Authors' contribution MK designed the study and wrote the first draft of the paper. All authors conducted fieldwork, collected and analysed the data. All authors revised and approved the final version of the paper. Acknowledgements
d
Our research was financially supported by the University of Ło z, statutory activity 511-1; grant number B1711000000123.01. We thank the editors and anonymous reviewers for their constructive comments, which helped us to improve the manuscript. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.actao.2017.09.004. References Ashcroft, M.B., 2010. Identifying refugia from climate change. J. Biogeogr. 37, 1407e1413. Babi
c V, Gali
c Z, Rakonjac L et al. (2010). Microclimate conditions in the stands of sessile oak on acid brown and lessive acid brown soils in Fruska Gora. Proceedings of the First Serbian Forestry Congress “Future with forests”. 11th-13th November 2010. Belgrade. Bakkenes, M., Alkemade, R.M., Ihle, F., et al., 2002. Assessing effects of forecasted climate change on the diversity and distribution of European higher plants for 2050. Glob. Change Biol. 8, 390e407. Birks, H.J.B., Willis, K.J., 2008. Alpines, trees, and refugia in Europe. Plant Ecol. Divers. 1, 147e160.
ski, M., 2011. Changing land use in recent decades and its Bomanowska, A., Kiedrzyn impact on plant cover in agricultural and forest landscapes in Poland. Folia Biol. Oecol. 7, 5e26. Braun-Blanquet, J., 1964. Plant Sociology, Broad Vegetation Science. 3 Aufl Springer, Wienn. Chytrý, M., 1997. Thermophilous oak forests in the Czech Republic: syntaxonomical revision of the Quercetalia pubescenti-petraeae. Folia Geobot. Phytotaxon. 32, 221e258.
dlo, J., 1997. Tilia-dominated calcicolous forests in the Czech Republic Chytrý, M., Sa from a central European perspective. Ann. Bot. (Roma) 55, 105e126. Chytrý, M., Danihelka, J., Hors
ak, M., et al., 2010. Modern analogues from the Southern Urals provide insights into biodiversity change in the early Holocene forests of Central Europe. J. Biogeogr. 37 (4), 767e780. Climate-Data.org (2016). pl.Climate-Data.org Available in 2016. Compton, J.A., Hedderson, T.A.G., 1997. A morphometric analysis of the Cimicifuga foetida L. complex (Ranunculaceae). Botan. J. Linn. Soc. 123, 1e23. Compton, J.A., Culham, A., Jury, S.L., 1998. Reclassification of Actaea to include Cimicifuga and Souliea (Ranunculaceae): phylogeny inferred from morphology, nrDNA ITS, and cpDNA trnL-F sequence variation. Taxon 47, 593e634. Davis, B.A.S., Brewer, S., Stevenson, A.C., Guiot, J., Data Contributors, 2003. The temperature of Europe during the Holocene reconstructed from pollen data.

8


ski et al. / Acta Oecologica 85 (2017) 1e8 M. Kiedrzyn

Quat. Sci. Rev. 22, 1701e1716.
rski, M., 1990. Warunki siedliskowe kateny ekosystemo
w le
snych na WysocDego
lnym uwzgle˛ dnieniem dynamiki wodno-troficznych zy
znie Rawskiej, ze szczego wła
sciwo
sci gleb. Geogr. Doc. 5e6, 1e206 (in Polish with an English summary). Ellenberg, H., 1988. Vegetation Ecology of Central Europe, 4nd ed. Cambridge University Press, Cambridge. Ellenberg, H., Weber, H.E., Düll, R., et al., 1992. Zeigerwerte von Pflanzen in Mitteleuropa. Ed 2. Scr. Geobot. 18, 1e258.
ski, M., Zielin
ska, K.M., Rewicz, A., 2014. The relationship beGrzyl, A., Kiedrzyn tween climatic conditions and generative reproduction of a lowland population of Pulsatilla vernalis: the last breath of a relict plant or a fluctuating cycle of regeneration? Plant Ecol. 215, 457e466. Hampe, A., Jump, A.S., 2011. Climate relicts: past, present, future. Annu. Rev. Ecol. Evol. Syst. 42, 313e333.
dl, R., Kopecký, M., Kom
He arek, J., 2010. Half a century of succession in a temperate oakwood: from species-rich community to mesic forest. Divers. Distrib. 16, 267e276. Jakubowska-Gabara, J., 1996. Decline of Potentillo albae-Quercetum Libb. 1933 phytocoenoses in Poland. Vegetatio 124, 45e59. Atlas Florae Europaeae. In: Jalas, J., Suominen, J. (Eds.), 1989. Distribution of Vascular Plants in Europe. 8. Nymphaeaceae to Ranunculaceae. The Committee for Mapping the Flora of Europe & Societas Biologica Fennica Vanamo, Helsinki. Jutrzenka-Trzebiatowski, A., Dziedzic, J., 1998. Vascular plants of the Jar Brynicy nature reserve. Nat. Sci. 1, 105e125. Kajtoch, Ł., Cie
slak, E., Varga, Z., Paul, W., Mazur, M.A., Sramko, G., Kubisz, D., 2016. Phylogeographic patterns of steppe species in Eastern Central Europe: a review and the implications for conservation. Biodivers. Cons. 25 (12), 2309e2339. Keppel, G., Van Niel, K.P., Wardell-Johnson, et al., 2012. Refugia: identifying and understanding safe havens for biodiversity under climate change. Glob. Ecol. Biogeogr. 21, 393e404. Keppel, G., Mokany, K., Wardell-Johnson, G.W., et al., 2015. The capacity of refugia for conservation planning under climate change. Front. Ecol. Environ. 13, 106e112.
ski M (2009). Vegetation succession tendencies in abandoned limestone Kiedrzyn and sandstone quarries in the Polish Uplands. 52nd International Symposium IAVS “Vegetation Processes and Human Impact in Changing World”. Chania, Crete. May 30th-4th June 2009. p: 187.
ski, M., Andrzejewski, H., 2012. Cimicifuga europaea Schipcz. pluskwica Kiedrzyn
dztwa europejska. In: Olaczek, R. (Ed.), Czerwona Ksie˛ ga Ro
slin Wojewo
dzkiego. Zagrozone _ _ Ło ro
sliny naczyniowe. Zagrozone zbiorowiska ro
slinne.
d

d

d
Botanical Garden in Ło z. University of Ło z, Ło z, pp. 64e65 (in Polish).
ski, M., Jakubowska-Gabara, J., 2014. The detection of thermophilous forest Kiedrzyn hotspots in Poland using geostatistical interpolation of plant richness. Acta Soc. Bot. Pol. 83 (3), 183e189.
ski, M., Zielin
ska, K.M., Kiedrzyn
ska, E., Jakubowska-Gabara, J., 2015. Kiedrzyn Regional climate and geology affecting habitat availability for a relict plant in a flat landscape: the case of Festuca amethystina L. in Poland. Plant Ecol. Divers. 8 (3), 331e341.
ski, M., Kurowski, J.K., Kiedrzyn
ska, E., Maciejewski, P., 2016. Exceptionally Kiedrzyn tall individuals in a relict population of Actaea europaea (Schipcz.) J. Compton against the species of the Euro-Asiatic Cimicifuga section. Pol. J. Environ. Stud. 25 (2), 889e893.
ski, M., Zielin
ska, K.M., Kiedrzyn
ska, E., Rewicz, A., 2017. Refugial debate: Kiedrzyn on small sites according to their function and capacity. Evol. Ecol. http:// dx.doi.org/10.1007/s10682-017-9913-4. Marks, L., Ber, A., Gogołek, W., et al. (Eds.), 2006. Geological Map of Poland, Scale 1: 500000. The Polish Geological Institute, Warsaw. Martynenko, V.B., Yamalov, S.M., Zhigunov, OYu., Filinov, A.A., 2005. Rastitel’nost’ Gosudarstvennogo Prirodnogo Zapovednika Shul’gan-Tash (Vegetation of the State Nature Reserve Shul’gan-Tash). Gilem, Ufa (in Russian).
ska, E., 1973. Charakterystyka i zro
znicowanie _ Matuszkiewicz, W., Roo-Zielin przestrzenne wła
sciwo
sci gleb w fitocenozie
swietlistej da˛ browy (Potentillo albae-Quercetum trollietosum) w Puszczy Białowieskiej. Phytocoenosis 2.3,

221e248 (in Polish with an English summary). €ger, E., Weinert, E., 1965. Vergleichende Chorologie der ZentralMeusel, H., Ja €ischen Flora. G Fischer, Jena. europa Moritz, C., Agudo, R., 2013. The future of species under climate change: resilience or decline? Science 341, 504e508.  Available in 2013. www.nature.portal.cz Agentura ochrany prírody a krajiny CR, 
Parusov
a, R., Mare ckov
a, L., Kapler, A., Majeský, L., Farkas, T., Indreica, A., Saf arov
a, L., Kitner, M., 2016. Adenophora liliifolia: condition of its populations in Central Europe. Acta Biol. Cracoviensia Ser. Bot. 58 (2), 83e105. Popov, G.V., 1980. Lesa Bashkirii (Forests of Bashkiria). Bashkirskoe Knizhnoe Izdate’lstvo, Ufa (in Russian). Puhe, J., Ulrich, B., 2001. Ecological Studies Series. Global Climate Change and Human Impacts on Forest Ecosystems: Development, Present Situation, and Future Trends in Central Europe, vol. 143. Springer-Verlag, Berlin, Heidelberg, New York. Ralska-Jasiewiczowa, M., Nalepka, D., Goslar, T., 2003. Some problems of forest transformation at the transition to the oligocratic/Homo sapiens phase of the Holocene interglacial in northern lowlands of central Europe. Veg. Hist. Archaeobotany 12, 233e247.
enz, S., Penas, A., et al., 2010. Computerized Bioclimatic Rivas-Martínez, S., Rivas-Sa Maps of the World: Bioclimate of Europe and Mideast. Draf Map Series of April 2010. http://www.globalbioclimatics.org/form/maps.htm. Role cek, J., 2007. Vegetation of Subcontinental Oak Forests in Central and Eastern Europe. PhD thesis. Masaryk University, Brno (in Chech with an English summary).
dlo, J., Chytrý, M., Pysek, P., 2007. Regional species pools of vascular plants in Sa habitats of the Czech Republic. Preslia 79, 303e321. Skov, F., Svenning, J.-C., 2004. Potential impact of climatic change on the distribution of forest herbs in Europe. Ecography 27, 366e380.  Smilauer, P., Leps, J., 2014. Multivariate Analysis of Ecological Data Using Canoco, 5nd ed. Cambridge University Press, Cambridge. StatSoft Inc, 2011. STATISTICA (Data Analysis Software System), version 10. www. statsoft.com. (Accessed 10 December 2013). Stewart, J.R., Lister, A.M., Barnes, I., et al., 2010. Refugia revisited: individualistic responses of species in space and time. Proc. R. Soc. B Biol. Sci. 227, 661e671. Szymura, T.H., Szymura, M., 2011. Soil properties and light availability determine species richness and vegetation diversity in an overgrown coppice oak stand. Pol. J. Ecol. 59 (3), 523e533. www.theplantlist.org version 1.1, September 2013, Available in 2017. Theurillat, J.P., Guisan, A., 2001. Potential impact of climate change on vegetation in the European Alps: a review. Clim. Change 50, 77e109. Thuiller, W., Lavorel, S., Araújo, M.B., et al., 2005. Climate change threats plant diversity in Europe. Proc. Natl. Acad. Sci. 102, 8245e8250. Tichý, L., Holt, J., 2006. Juice e Program for Management, Classification of Ecological Data. Program Manual. Vegetation Science Group, Masaryk University, Brno. van Andel, T.H., Tzedakis, P.C., 1996. Palaeolithic landscapes of Europe and Environs, 150,000e25,000 years ago. Quat. Sci. Rev. 15, 481e500. Vera, F.W.M., 2000. Grazing Ecology and Forest History. Cabi Publishing, Oxfordshire. Wallingford. Willis, K.J., van Andel, T.H., 2004. Trees or no trees? The environments of central and eastern Europe during the Last Glaciation. Quat. Sci. Rev. 23, 2369e2387. Woolbright, S.A., Whitham, T.G., Gehring, C.A., et al., 2014. Climate relicts and their associated communities as natural ecology and evolution laboratories. Trends Ecol. Evol. 29, 406e416. Young, A.G., Boyle, T., Brown, T., 1996. The population genetic consequences of habitat fragmentation for plants. Trends Ecol. Evol. 11, 413e418. Zaja˛ c, A., Zaja˛ c, M. (Eds.), 2001. Distribution Atlas of Vascular Plants in Poland. Laboratory of Computer Chorology, Institute of Botany, Jagiellonian University, Cracow.
ska, K.M., Kiedrzyn
ski, M., Grzyl, A., Rewicz, A., 2016. Forest roadsides harbour Zielin less competitive habitats for a relict mountain plant (Pulsatilla vernalis) in lowlands. Sci. Rep. 6, 31913.