Specific Vicariance of Two Primeval Lowland Forest Lichen Indicators

Many articles underline that tree species diversity and composition is one of the fundamental factors explaining epiphytic species distribution within forests (see Mežaka et al. 2008; Király et al. 2013). However, in this study it turned out that the studied plots could not be differentiated in terms of tree composition (Fig. 3 and Online Resource 2). Neither was there any significant difference in tree density. Thus it can be assumed that canopy closure and light availability were comparable in all of the plots (Fig. 4). The age of a stand is another key factor responsible for lichen biodiversity and composition (Fritz et al. 2008a, b; Nascimbene et al. 2010b). The oak-hornbeam forest stands inhabited by C. candelaris were generally older than those populated by V. hemisphaerica (Fig. 4). Additionally, the general age of forest stands was the only factor to positively affect the abundance of C. candelaris in the plots. It might simply be assumed that C. candelaris is attached to older forests. However, this species also appeared in the neighbouring 120-year-old mixed managed forests, provided that big oaks were present (Kubiak et al. 2016). Moreover, only the mean diameter of oaks in the plots, which was the essential host tree for C. candelaris (Fig. 5), significantly correlated—at least in this study (Online Resource 3a)—with the age of forest stands. Aged oaks, which are more likely to remain in old and unaffected oak-hornbeam stands, provide the microhabitat space necessary for C. candelaris. Varicellaria hemisphaerica has been frequently recorded in deciduous forests of uneven age, including those over two hundred years old (e.g., Fritz et al. 2008b). Therefore, we assume that the estimated fifty-year difference in the general age of current stands is not a direct and key factor regulating the occurrence of these species. It is very likely that the qualitative and age structure of the forest complex considered as a whole on a large-landscape scale has a greater influence on distribution patterns than the stand parameters directly characterising the examined plots. Although the coverage participation of forest stands over a hundred years old in the N and P complexes (approximately a third of the total) is somewhat higher than in B and R (less than one-fifth), the total participation of deciduous trees in the first two complexes falls within the vicinity of 10% (see Fig. 1). We believe that the vast total area of deciduous forest consistent with the habitat, along with its relatively low fragmentation, is of great importance for maintenance of V. hemisphaerica. On the other hand, C. candelaris showed no sensitivity to this general factor and even tended to colonise old mixed forests of anthropogenic origin (see Kubiak et al. 2016) or natural forests under strong human influence (Motiejūnaitė 2009).

The occurrence of the examined lichen indicators is highly dependent on the geographical location of the forest complexes in which they grow. A relatively slight difference in location may be associated with quite different climate parameters. The studied complexes lie at similar distances from the coast of the Baltic Sea (about 100–150 km) and are not subject to the direct influence of a maritime climate. However, considering the study area, C. candelaris clearly prefers south-western complexes (the N and P forests), V. hemisphaerica north-eastern (the B and R forests). The complexes comfortable for V. hemisphaerica are located in an area with significantly higher relative humidity and precipitation as well as lower mean temperature compared to those preferred by C. candelaris (Fig. 2). The preference of C. candelaris for drier habitats and the attachment of V. hemisphaerica to moderately moist habitats are suggested based on an overview of the data in the literature (Fabiszewski and Szczepańska 2010). According to Wirth (2010), both lichens are characterised by similar ecological indicator values, although the moisture value is somewhat higher for V. hemisphaerica. The presence of C. candelaris in mixed forests with a high proportion of pine (see Kubiak et al. 2016) also indirectly suggests its low sensitivity to dehumidify of habitat (see also von Arx et al. 2012). The highest annual rainfalls in north-eastern Poland, which regularly exceed 700 mm, occur in the region of the R forest. These values are typical for the coastal zone of western Poland, where V. hemisphaerica is characterised by the highest frequency in the lowland part of the country within the beech distribution range (Fałtynowicz 1992). Mean annual precipitation is the most effective ecological predictor of the occurrence and diversity of many lichens (Svoboda et al. 2010; Merinero et al. 2014). In central Spain, V. hemisphaerica has been classified among the group of lichens very sensitive to the ‘edge of forest effect’ and strongly prefers the forest interior (Belinchón et al. 2007; Aragón et al. 2015). On the other hand, in Western European areas with very high humidity, this species appears in open habitats (Tønsberg 1992; Chambers et al. 2009). The dependence of V. hemisphaerica on climatic conditions and sufficient moisture of habitat may explain its low prevalence in the Białowieża Forest (Cieśliński 2003), one of the best preserved natural lowland forests in Europe, where suitable host trees certainly exist, but where annual rainfall and relative humidity are not as high as in the B and R forest complexes (Faliński 1986). Understanding the climatic sensitivity of species is a central theme in biodiversity conservation. Climate and response to climate change is one component among a large number of drivers which interact to control the occurrence of lichens (Lisewski and Ellis 2011).

Vertical distribution patterns of lichens have been usually investigated in mountain areas (Çobanoğlu and Sevgi 2009; Nascimbene and Marini 2015); the significance of elevation for lichen occurrence in the lowlands is rarely taken into consideration. However, the topography of some lowland areas may be characterised by highly diversified relief and great differences in elevation which often induce local climatic fluctuations. The average height above sea level of the B and R forests is higher by about 50 m than that of the N and P forests (Fig. 4). The elevation forces the convection of polar maritime air masses on the slopes of moraine hills and, furthermore, promotes condensation of water vapour in the air, consequently increasing the amount of rainfall (Huggett and Cheesman 2002). In the end, elevation turned out to be the only factor (rather than variables describing the general structure of stands) positively and significantly affecting V. hemisphaerica abundance within the plots. This is another point indicating the high moisture requirements of the species.

The presence of large old trees in the forest complex undoubtedly favours the occurrence of C. candelaris, which demonstrates strong dependence on oak trunks (Fig. 5 and Online Resource 3B). The age and bark structure of trees are among the most important factors determining the occurrence of certain epiphytes (Uliczka and Angelstam 1999; Johansson et al. 2007; Fritz et al. 2008a; Marmor et al. 2011). The average age of forest stands for plots with C. candelaris amounted to nearly 200 years; hence old and massive oaks have been preserved in large numbers in these forest complexes. The stands (including oaks) in the B and R forests are younger (Fig. 4); this may be one of the important causes for the negligible presence of C. candelaris in these complexes (see also Zalewska and Fałtynowicz 2004; Zalewska 2012; Forest Data Bank 2016). The same applies to maples, though C. candelaris shows a lesser preference for the trunks of this tree. The greater total sum of tree diameters and mean diameter of oaks and maples in the plots with C. candelaris compared to those with V. hemisphaerica (Fig. 4 and Online Resource 3B) suggests that the former species requires the provision by suitable host trees of a sufficiently large microhabitat surface. The positive impact on lichen species richness exerted by a high density of oaks in the immediate area in a larger landscape has been demonstrated (Ranius et al. 2008a, b; Paltto et al. 2010). Moreover, old oaks constitute almost the only harbour for C. candelaris in old managed forests planted in habitats typical of temperate deciduous forest (Kubiak et al. 2016). Numerous sources of lichen propagules and their proximity to suitable trees increase the probability that other trees will be colonised by epiphytic lichens (Johansson et al. 2009). In natural forests (e.g., the Białowieża Forest), where C. candelaris occurs abundantly, it sometimes grows on different trees (Cieśliński and Tobolewski 1988; Cieśliński 2003). Perhaps large and firmly established populations show greater tolerance in regard to host trees.

Varicellaria hemisphaerica is characterised by a much lower level of specialisation regarding species and age of host trees. Even though the number of plots with C. candelaris and V. hemisphaerica was the same, the latter was recorded on twice as many trees. V. hemisphaerica was recorded on almost all species of deciduous trees, with the exception of birch. However, hornbeam and oak constituted its main preferences (Fig. 5) and hornbeams with a large diameter were favoured by individuals (Online Resource 3b). In other deciduous forests, the thalli of V. hemisphaerica also readily overgrow beech bark, and this lichen can also be encountered on alders, hazels, and rowans (Fałtynowicz 1992; Cieśliński 2003). The bark of oaks, hornbeams, and lindens is fairly acidic (Barkman 1958; Jüriado et al. 2009) but the structure of these kinds of bark differs at various stages of the trees’ lives. It seems that a porous periderm structure, which appears only in old hornbeams, is an important factor for this lichen (Mežaka et al. 2008). In the Baltic countries, at the extremes of beech and hornbeam zone coverage, V. hemisphaerica inhabits mainly oaks (Prigodina Lukošienė and Naujalis 2006; Motiejūnaitė et al. 2008). The wide tolerance of V. hemisphaerica in relation to host trees may explain its frequency (twice that of C. candelaris) in the studied plots.

Although our target species were C. candelaris and V. hemisphaerica, the study also concerned more than 100 accompanying lichen species. It can be stated that a large part of the lichen epiphytic biota characteristic of this type of forest community has been found to be directly associated with these indicators (cf. Cieśliński et al. 1995; Mežaka et al. 2008; Jüriado et al. 2009; Motiejūnaitė and Prigodina Lukošienė 2010; Svoboda et al. 2010, 2011). A slightly higher number of species recorded together with V. hemisphaerica results from the greater diversity of main host trees (Table 1 and Fig. 5). Both indicators co-create specific assemblages of epiphytic lichens (Fig. 6 and Online Resource 4) characterised by a high number of exclusive species (Table 1). These assemblages are similar in terms of species diversity but differ significantly in terms of species composition. Moreover, many lichens with high conservation value are integrated in both sets of lichens. Among the red-listed and endangered lichens in Poland (according to Cieśliński et al. 2006), Arthonia byssacea and Calicium viride (for the C. candelaris assemblage) and Arthonia vinosa, Pertusaria coronata, P. pertusa, Pyrenula nitida, and P. nitidella (for the V. hemisphaerica assemblage) were recorded most often. The great distinctness of the C. candelaris and V. hemisphaerica assemblages is symptomatic and indicates that their exclusive members exhibit similar habitat requirements. In addition, the intimacy of many accompanying species is frequently manifested in relation not only to C. candelaris and V. hemisphaerica but also to their main host trees (Fig. 7). Environmental evaluation should consider not only the separate occurrence of individual species but also the occurrence of sets of species forming specific communities. Assemblages of certain lichens are characterised by a high level of repeatability, whereas the appearance of individual species may be coincidental and insufficient for a proper diagnosis (see also Rola and Osyczka 2014).

Oak-hornbeam forests constitute the primary type of potential vegetation over large lowland areas of Europe, but at the same time are, due to agricultural expansion and the cultivation of wood-productive pines and spruces, the most endangered (Matuszkiewicz 2008). Therefore, this forest community has been included in the ecological network of special protected areas in Europe known as Natura 2000. One of the main objectives of this project (Council Directive 1992) is to preserve or restore valuable natural habitats. In practice, specifying the present condition of the habitat and determining the size of the protective area to be established for successful maintenance of the habitat are not simple tasks. Well-chosen lichen indicators can be of great assistance in such endeavours. Notwithstanding, the objective specification of lichen habitat requirements should be accomplished first, in order to define their actual indicative application.

Certain lichens are readily used to ascribe conservational value to old forest stands or natural woodlands (Nordén and Appelqvist 2001). Based on species composition, many biotas closely related to specific forest communities have been reported, and numerous lichens have been appointed as indicators of ‘ecological continuity’ (e.g., Rose 1974), ‘primeval (virgin) forest’ (e.g., Cieśliński et al. 1996), or ‘lowland old-growth forest’ (Czyżewska and Cieśliński 2003; Motiejūnaitė et al. 2004). Imprecise definitions of these concepts, ambiguous criteria for the selection of such indicators, and, consequently, interpretations of environmental information potentially provided by the occurrence of these lichens, often give rise to dispute (e.g., Nordén and Appelqvist 2001; Rolstad et al. 2002; Nordén et al. 2014). Specific local habitat parameters, the period required for the effective colonisation of suitable substrates, and natural disturbances of forest structure may in fact constitute key factors responsible for lichen biota development and the presence or absence of particular species (Kalwij et al. 2005). Our research showed that identification of potential habitat factors affecting the occurrence and abundance of particular indicators in the greatest possible detail is a desirable step towards the verification of their true bioindicative usefulness, especially in the context of a regional environmental evaluation.

Varicellaria hemisphaerica is not host-specific and readily inhabits trunks of various deciduous trees. A sufficiently high level of moisture in the habitat is essential for the species, which prefers forests with a high proportion of deciduous trees on a large-landscape scale. Habitat continuity on the level of the local landscape is more important for this species than the current age and structure of forest stands; thus, this lichen appears to be a good indicator of the ecological continuity of regional varieties of oak-hornbeam forest (see also Kubiak 2011, Kubiak and Łubek 2016). The opposite pattern is demonstrated by Chrysothrix candelaris. Among the many factors that determine the specificity of a given habitat, its occurrence basically depends on only one, i.e., the microhabitat space provided by old oaks. The lack of such trees in a given stand may seriously inhibit the lichen. Thus, its role as a general indicator of forest continuity in the context of a whole habitat is very limited.

Regardless of the indicative value, our study proved that localities within oak-hornbeam forests inhabited by both C. candelaris and V. hemisphaerica deserve special protection status. Many exclusive and endangered lichens with similar habitat requirements are associated with these two species (Table 1). Protecting their habitats may indirectly ensure the effective protection of many other lichens that make up coherent and stable epiphytic biotas. Moreover, the spontaneous restoration of deciduous forest areas consistent with the habitat, along with the associated lichen biota, is possible within a relatively short period. The occurrences of C. candelaris individuals on oaks in human-transformed mixed forests should be regarded as a positive phenomenon that may indicate that the process of natural regeneration is underway (see also Kubiak et al. 2016). Dense populations of C. candelaris and V. hemisphaerica may be a useful environmental tool for the designation of protected areas as ‘forests possessing unique environmental value’, according to the criteria of the High Conservation Value Forests programme HCVF (WWW 2007).