Introduction

The Saale–Unstrut area is located in the Burgenlandkreis (a district in South Saxony-Anhalt, Germany). The high density of edifices (churches, cathedrals, monasteries, historic industrial buildings and others) in the river valleys of the Unstrut and Saale creates the appearance of a closed “cultural landscape corridor” (cf. Hoppert et al. 2018). Most historic stone monuments were constructed by locally available Triassic sandstones (“Buntsandstein”) and limestone (“Muschelkalk”). Both types are highly variable with respect to, e.g., grain size, cementation, porosity and hence general weathering susceptibility (Stück et al. 2013). Physical/chemical and biological weathering phenomena attack the stone surface simultaneously and depend on each other. Slightly weathered dimension stone, in particular, is susceptible to colonization by endolithic organisms like unicellular eukaryotic algae, bacteria, mosses and lichens (Gaylarde et al. 2003). Typical initial effects of microbiological colonization are discolorations of building stones, which mainly affect the appearance of architectural decoration (Gorbushina et al. 1993; Hallmann et al. 2011a, b). Progression of weathering leads to degradation of the building stone surface in various ways, which makes it difficult to attribute advanced decay phenomena to a single initial event such as, e.g., salt splitting or microbial growth (e.g., Stück et al. 2013; Hallmann et al. 2013a, b).

Generally, microbial endoliths penetrate stone just some µm up to few mm. However, depending on the pore size, stones may be colonized several centimeters below the surface in particular by (lichen) fungi and mosses (Hallmann et al. 2014a). Microorganisms and cryptogams benefit from the microhabitat inside the stone. Adverse environmental conditions, like high radiation, rapid desiccation, effect of extreme temperatures or grazing are reduced in this ecological niche (Griffin et al. 1991; Hoppert et al. 2004; Gorbushina 2007).

Algae, dematiaceous fungi (cf. Hallmann et al. 2011a, 2013b), and lichens (cf. González-Gómez et al. 2018) destain and deteriorate surfaces in advanced stages of colonization. Assessment of initial microbial colonization by molecular methods may be helpful to decide on further conservation measures.

The aim of this study is to give an inventory of stone associated (micro-)organisms on sandstone and limestone which may help to give, along with other physical methods of assessment of stone decay (Stück et al. 2013, 2018), a state-of-the-art report on monument degradation.

Materials and methods

Sampling site and preparation of samples

Samples from limestone lithologies (Muschelkalk, “Schaumkalk”) originate from Saaleck Castle/Bad Kösen (Saaleck rampart: 51.109466 N, 11.701846 E, Saaleck wall base: 51.109526N, 11.701608E; cf. Fig. 1a, b). Samples of Buntsandstein (“Hardegsen-formation”) were taken in early June 2012 from the location Blütengrund near Großjena/Naumburg (Saale) (51.183035N, 11.788119 E and 51.177971N, 11.793808E; Fig. 1c). Another sampling site close to this location (51.178110N, 11.793193E), the “Stone Album” (“Steinernes Bilderbuch”), consists of sandstone from the same formation (Fig. 1d).

Fig. 1
figure 1

Sampling sites in the Saale–Unstrut area. a Rampart of Saaleck Castle. b Section of the wall base of the Saaleck Castle with outcropping rock (Schaumkalk). c Outcropping rock (Buntsandstein, Hardegsen-formation, Großjena). d Outcropping rock, “Stone Album” near Großjena (photograph provided by H. Stück); white dots in b, c mark selected sampling points

Full size image

Generally, sampling sites were allocated according to accessibility, compatibility of sampling spots with respect to monument preservation regulations and apparent signs of microbial colonization (discoloration, signs of biogenic weathering). In total, 60 sampling spots were selected. A sampling spot is defined as a small area of approximately 4 cm2. In this area, samples were scratched from the surface with a sterile scalpel and were immediately transferred to sterile plastic containers. Among these 60 original samples, 23 could be further processed (19 environmental samples, 4 cultures, cf. Tables 1, 2, 3). Seven samples were taken from the location Blütengrund/Großjena (Buntsandstein, Hardegesen-Folge). The rock face is exactly W-exposed. The rock surface showed common signs of backweathering (cf. Fig. 1c; some sampling spots are marked by white dots), and spots were selected in a way that different microtopographies (resulting from backweathering) were included, but spots with apparent accumulation of soil, mosses and lichens (horizontal surfaces and clefts) were excluded. For surface samples, care was taken that no contamination from below the surface was collected (for sampling of depth profiles see below). Another two samples were taken from the site “Stone Album”, from a vertical rock face in SSW exposition. Here, sampling was restricted due to monument protection regulations.

Table 1 Algal phylotypes retrieved from all sampling spots on limestone and sandstone as well as from crude cultures
Full size table
Table 2 Phylotypes of protozoa, lichens and fungi, retrieved from all sampling spots on limestone and sandstone as well as from crude cultures
Full size table
Table 3 Algal, fungal and other phylotypes retrieved from surface samples and from depths of 0.5 cm beneath the surface as indicated (W = west-exposed; S = south-exposed)
Full size table

Saaleck was sampled on two exactly S-exposed and W-exposed vertical rock faces (cf. Fig. 1a, b). However, microtopography and hence exposition were again fluctuating according to irregularities of the natural rock face or the roughly trimmed dimension stone. In total, six sampling spots were selected from those sampling sites (three spots from S-exposed wall base, three spots from W-exposed rampart).

For analysis of depth profiles, one sampling spot from Großjena and from Saaleck was selected, respectively. Both spots exhibited microtopographies with W- and S-exposed surfaces. Within a distance of 10 cm, both W- and S-exposed surfaces were sampled separately as described above. Then, the whole rocks were removed from the site and cracked with a chisel under sterile laboratory conditions within several hours after sampling. Samples were then taken, as described above, from a spot below the surface as described in Hallmann et al. (2014a).

For preparation of crude cultures, aliquots of four selected samples (two from each sampling site, either Saaleck or Großjena) were suspended in 20 ml 3N BBM+V medium and Z Medium in 100-ml Erlenmeyer flasks (Starr and Zeikus 1993; Watanabe and Nozaki 1994), respectively. After inoculation, cultures were incubated at constant temperature of 18 °C. White fluorescent illumination with an intensity of 25 µmol photons m−2 s−1 was applied for four weeks, while the light:dark cycle was set to 14:10 h.

DNA extraction

Extraction of genomic DNA from stone samples was performed using the DNeasy PowerSoil DNA isolation kit (Qiagen, Hilden, Germany) according to the manufacturer’s instruction. Successful DNA extraction (see below) required up to 250 mg of crushed samples. Samples were further processed as described in Hallmann et al. (2014a).

For preparation of crude cultures, an aliquot of 0.5 ml (0.25 ml of cultures from a sampling spot in BBM + V medium, and Z medium, respectively) was transferred to 2 ml beat-beating tubes and mixed with equivalent amounts of acid-washed glass beads (120–200 µm and 425–600 µm in diameter; Sigma-Aldrich, ST. Louis, MO, USA). These tubes were treated for 30 s at 5000 rpm in a Minibeadbeater (Biospec, Barlesville, OK, USA). DNA was extracted using the Invisorb® Spin Plant Mini Kit (Stratec Molecular, Berlin, Germany), following the manufacturer’s instructions. Sampling on a 1% (w/v) agarose gel confirmed successful DNA extraction. Isolated DNA was stored at −20 °C until further processing.

Polymerase chain reaction (PCR) amplification

PCR amplification was performed for isolated biofilm DNA using eukaryote-specific primer combinations for 18S rRNA gene, 20F (5′ GTAGTCATATGCTTGTCTC 3′; Thüs et al. 2011) and 18L (5′ CACCTACGGAAACCTTGTTACGACTT 3′; Hamby et al. 1988). Templates comprised approximately 10–100 ng of DNA. Amplification reaction mixture (25 µl) contained each dNTP at a concentration of 0.1 mM, 5 µl of 10× reaction buffer, 2 mM MgCl2, each primer at a concentration of 0.2 µM, 2 U of Taq DNA polymerase (Bioline, Luckenwalde, Germany) and 4% (v/v) dimethyl sulfoxide (DMSO)-solution. PCR was performed in a thermocycler TProfessional Basic (Biometra, Göttingen, Germany) using the following program for the primer set 20F/18L: initial denaturation at 95 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 1 min, annealing at 50 °C for 1 min, extension at 72 °C for 3 min and final extension at 72 °C for 10 min. The PCR products were purified using the InvisorbR Spin PCRapid Kit (Stratec molecular). Aliquots of 2 µl of purified amplicons were analyzed by electrophoresis on a 1% (w/v) agarose gel to check for amplification.

18S rRNA gene cloning and sequencing

Cloning was carried out with the TOPO TA cloning kit (Life technologies, Carlsbad, CA, USA) using TOP 10 chemically competent One Shot—Escherichia coli cells (Life technologies), as supplied by the manufacturer. All eukaryotic clones were sequenced with the 18S rRNA gene standard sequencing primer 895R (5′AAATCCAAGAATTTCACCTC 3′) resulting in partial sequences including the hypervariable regions V2-V4 (Hodač et al. 2012). Sequencing reactions were performed by Macrogen Inc. (Seoul, South Korea).

Sequence analysis and phylogeny

Resulting sequences were manually corrected using the sequence analysis program SeqAssem (Hepperle 2004). These sequences were analyzed by BLASTn with NCBI database (Altschul et al. 1990, http://www.ncbi.nlm.nih.gov/). Analyzed sequences and reference sequences were imported into the ARB program (Ludwig et al. 2004, http://www.arb-home.de). To determine phylogenetic affiliations, relevant sequences were aligned with the homologous eukaryotic 18S rRNA gene sequences using the automatic alignment tool of the ARB program package. Potential chimeras were checked with Bellerophon (Huber et al. 2004). In addition, the first and the last 300 bp of putative chimeras were compared with similar rRNA gene sequences in NCBI and excluded from the dataset. After chimera check, 568 clones could be retrieved. These clones were assigned to 94 phylotypes (several eukaryote taxa), grouped at a similarity of 97% or higher to the closest related sequence retrieved by BLASTn search (Tables 1, 2, 3).

A phylogenetic tree (supplement Fig. 1) was constructed with representative full length sequences of algal phylotypes using the RAxML search algorithm for maximum likelihood (ML; Stamatakis et al. 2008), using the GTR + Γ + I model. The confidence of the tree topologies was tested by bootstrap analysis implemented in RAxML (100 replicates) and by Bayesian posterior probabilities (MB) using MrBayes 3.2 (Huelsenbeck and Ronquist 2001). Two parallel Markov chain Monte Carlo (MCMC) runs for one million generations each with one cold and three heated chains were conducted using the GTR + Γ + I model, with trees sampled every 100 generations. Sequence alignment was performed using MAFFT (Katoh and Toh 2008). The alignment consisted of 117 sequences with 1802 positions (738/534 variable/parsimony informative) in total. Sequences derived from 15 environmental clones, assigned to algal phylotypes (this study) and 97 isolates from culture collections as well as environmental clones from other studies (cf. Hallmann et al. 2013a, b).

Representative sequences were deposited in GenBank under the following accession numbers: MH807077–MH807091.

Results

Exemplary for many dimension stones in the Saale–Unstrut region, the Schaumkalk at the base of Saaleck castle (cf. Fig. 1a, b) and the Buntsandstein (Hardegsen-formation, Fig. 1c, d) were selected for assessment of microbial diversity on stone surfaces. Tables 1 and 2 show data from each sampling spot and from crude cultures. In Fig. 2, summarized data from most abundant phylotypes (more than one per spot) at a sampling site are shown (crude cultures are excluded here). In total, 94 phylotypes were identified, mainly belonging to algae and filamentous fungi (including lichen fungi). Other retrieved phylotypes could be assigned to protozoa, arthropods, mosses and ferns. Clearly, phylotypes and numbers of retrieved sequences per phylotype varied; many phylotypes were just detected once in a sample. In cultures inoculated with samples from sandstone and limestone surfaces, phylotypes were identified that were not retrieved from clone libraries of the environmental samples. In particular, freshwater algae and protozoa of diverse phylogenetic groups were retrieved.

Fig. 2
figure 2

Simplified heatmap displaying the number of retrieved clones, assigned to abundant phylotypes (sum over all sampling points; cf. Tables 1, 2) from two limestone (Saaleck) and two Buntsandstein (Großjena and Stone Album) sites

Full size image

Among algae directly retrieved from stone surfaces (Table 1; Fig. 2), differences between limestone and sandstone are mainly due to the Trebouxiophyceae clone QE59, which is quite abundant in limestone samples, but present in just one sandstone sampling spot. We retrieved a Phyllosiphon arisari (putatively the species was wrongly assigned to a sequence; cf. Hallmann et al. 2013a; Procházková and Neustupa 2016)-related phylotype from both sets of samples. Other algae were present in small numbers (Chlorella-, Stichococcus-, Pseudostichococcus-, Desmococcus-related).

With respect to (lichen) fungal phylotypes, (Table 2; Fig. 2), differences between sandstone and limestone are more obvious. Although only spots without visible lichen thalli were processed, samples from Saaleck (limestone) were dominated by phylotypes of lichen fungi, in particular Xanthoria elegans, and to a lesser extent, Texosporium sancti-jacobi. Both Saaleck sampling sites appear to be similar. Penicillium solitum-related phylotypes are high abundant at only one spot from Saaleck rampart wall (cf. Table 2). Penicillium solitum is a plant pathogenic fungus (Pitt et al. 1991). Others are related to insect- or other plant-associated fungi, like Cordyceps brongniartii (Shimazu et al. 1988) or Rhytidhysteron rufulum, Cryptococcus carnescens, Cladosporium bruhnei, Glyphium elatum, Pleospora herbarum and Phaeosphaeria nodorum (Takashima et al. 2003; Schubert et al. 2007; Hane et al. 2007; Woudenberg et al. 2017; Boehm et al. 2015; Chokpaiboon et al. 2016).

The fungal community on Buntsandstein is considerably different (Table 2). The Xanthoria elegans lichen fungus is missing, but on three spots, the Caloplaca demissa lichen fungus is abundant. Rhinocladiella and Knufia perforans are present in most of the sandstone sampling spots. Rhinocladiella is a typical rock-associated fungus (Sert et al. 2007; Hallmann et al. 2013b), but was also isolated from lichens (though it is not necessarily a lichen fungus, Harutyunyan et al. 2008). Knufia (Coniosporium) perforans has also been described as a rock-inhabiting fungus (Sterflinger et al. 1997). Also Capnodiales comprise plant and rock-associated phylotypes (Crous et al. 2009; Hallmann et al. 2013b). Arachnomyces canei has been described as a human pathogen (Gibas et al. 2002), but other members of the genus are rock-inhabiting fungi (Gueidan et al. 2008). Occasionally, we retrieved small animals (Oribatula tibialis) from the clone libraries, possibly due to the presence of eggs.

In summary, the calcicolous Xanthoria lichen fungus deserves attention due to its presence on most of the limestone surfaces, and Rhinocladiella, in particular, on Buntsandstein.

A small set of samples was taken to elucidate the inventory of endolithic organisms. Due to their microtopography (e.g., cliff formation, formation of edges due to backweathering and material loss), rock slope and surface exposition is variable. For comparison of organisms attached to a rock surface with a community at a depth of 0.5 cm below the surface, two surface samples (exposed roughly south and westwards in the field) were taken. The rock piece was removed, cracked and another sample from inside the rock was taken. A summary of the results is shown in Table 3.

From sampling site Großjena (sandstone), and Saaleck (limestone) phylotypes of algae (mainly Trebouxiophyceae), of mosses, and of fungi (mainly ascomycetes) were retrieved from surfaces, irrespective of their exposition. In samples taken at a depth of 0.5 cm, algae are missing almost completely. All other microorganisms belonged to plant-associated/plant-pathogenic fungi. From both sampling sites, tardigrade phylotypes were retrieved (Halobiotus/Macrobiotus).

Discussion

Specific microbial communities colonize all surfaces of rocks and dimension stone. Among eukaryotes, algae are the most important primary producers, but ascomycete fungi are also highly abundant (Gorbushina et al. 1993; Gorbushina 2007; Hallmann et al. 2011a, b, 2013a, b, 2014a, b, 2016). Among them, lichen fungi are one important group. Though no lichen thalli were visible on any of the sampled surfaces, phylotypes of lichen fungi and the lichen alga Trebouxia were abundant according to the analyzed clone libraries. Many sampling spots were dominated by either Xanthoria elegans or Caloplaca (Lecanora) demissa lichen fungi. The preferred substrata of the lichen fungi correspond to those of the well-developed lichen, i.e., either calcareous rocks for X. elegans or siliceous rocks for C. demissa, though both lichen species have broad ecological amplitudes (cf. Wirth 1995).

Generally, lichen thalli were abundant on limestone and sandstone surfaces in the area. At the sampling spots, however, no thallus structures (cf. Wirth 1995) were observed. In spite of this, phylotypes of lichen fungi (i.e., fungal species, known to be part of a lichen symbiosis) and the lichen alga Trebouxia were detected. This may account for the presence of lichen prothalli, invisible to the naked eye, which may develop to a visible lichen thallus during the following years (cf. Sanders 2014). The presence of the lichen fungus of Texosporium sancti-jacobi may account for a rather flexible life style of some symbiotic partners in lichens. Texosporium sancti-jacobi thalli were found rarely on Western North American semi desert soils (McCune and Rosentreter 1992; Riefner and Rosentreter 2004), but not in other areas. Hence it must be assumed that the lichen fungus (or a closely related genus) is much more abundant than the well-developed lichen. This assumption is reasonable, because ascomycete fungi develop a sexual reproductive stage (teleomorph) rather rarely (frequently it is unknown), though the inconspicuous mycelia of the asexually reproductive stage are abundant (Cannon and Kirk 2000). In visible lichen thalli, in contrast, the lichen fungus frequently reaches its teleomorph state—which is also the case for Texosporium (Tibell and v. Hofsten 1968). Hence, the presence of lichen fungal mycelia, extending on the surface of or inside a substratum should be considered (cf. Hawksworth 1988). As a consequence, the building stone may be affected by lichen fungi, even when visible thalli are absent.

Many clones affiliated to non-lichen fungi were also detected. The largest subset of the retrieved fungi (among them Glyphium elatum, Rhinocladiella sp. and Knufia sp. are most abundant) exhibits melanized cell walls. These dematiaceous fungi are phylogenetically diverse and typical for extreme, dry, sun-exposed habitats. Many of them were found to be involved in colonization and degradation of natural and artificial building material (Gorbushina et al. 1993; Gorbushina 2007). Generally, the fungi discolor surfaces, but also penetrate existing fissures and enhance chemical and physical weathering processes.

Among algal phylotypes, we retrieved mainly Trebuxiophyceae (cf. Table 1; Fig. 2). Many algae of this group are adapted to terrestrial habitats. Trebouxia, in particular, is the most important lichen alga and is putatively an obligate symbiont—in contrast to the fungal counterparts (cf. Amadijan 1988; Bates and Garcia-Pichel 2009). Another alga, “Phyllosiphon arisari” is also abundant and appears to be a common colonizer of dimension stone surfaces (cf. Hallmann et al. 2013a, b) and is mainly, but not exclusively, present on limestone surfaces. The diversity of trebouxiophyceaen vs. chlorophyceaen alga in our samples is supported by the phylogenetic tree (Fig. S1). Generally, Trebuxiophyceae may be considered as long-term sub-aeric rock colonizers, whereas the upcoming of other algal groups may be taken as an indicator for increasing moisture, or the presence of liquid water. Consequently, in liquid cultures, besides Trebouxiophyceae, also chlorophyceaen algae and protozoans were enriched. This feature has been repeatedly described and suggests the presence of a diaspore bank of algae and resting stages of protozoans that could become relevant when environmental conditions change, e.g., when the stone surface is moistened by rainfall (Hallmann et al. 2013a, b, 2014a, 2016).

Algal endoliths were first discovered in a zone of few millimeters beneath a surface of a rock in extreme dry and cold Antarctic valleys (Friedmann 1982). Colonization of rock substrata down to depths of several centimeters (Hallmann et al. 2014a; Cockell et al. 2017) or even hundreds of meters (e.g., Breuker et al. 2011) is also possible, though algae should be not expected. A 0.5-cm thick piece of the microcrystalline or amorphous rock absorbs all light (other than rocks consisting of rather large crystals, e.g., calcite); consequently photoautotrophic organisms were not retrieved from most of the samples in our study. Residual green algae in few sampling spots may be due to contamination from the stone surface, or transport by animal vectors. However, it is reasonable to assume that fungi penetrate the stone matrix, which is in particular true for porous rocks. As endoliths, fungal phylotypes do not necessarily belong to dematiaceous fungi, putatively because of the absence of stress by ultraviolet light. The presence of moss clones may be explained by the subterrestrial moss protonemata (Hallmann et al. 2014a).

Large animals such as tardigrades, which were exclusively endolithic in this study, indicated the presence of pores of several hundreds of µm in size. Tardigrades are omnivorous and feed on nematodes, moss, fungi or bacteria (Sánchez-Moreno et al. 2008; Schill et al. 2011). Their presence inside rocks may be taken as a strong signal for progressive degradation of a stone matrix.

Not all weathering phenomena cause damage in the sense of an actual loss of value. In many cases, discoloration and superficial colonization of dimension stone surfaces are of little relevance with respect to conservation of a monument. However, is certainly important to slow down the rapid weathering of the sandstone reliefs of the “Stone Album”, because of its uniqueness as a rock monument in Central Europe (cf. Stück et al. 2018; Hoppert et al. 2018). Though colonization by algae and fungi is often perceived just as an esthetic limitation, endolithic colonization by fungi and by animals indicates penetrable (open) pores, which inevitably leads to surface deterioration and material loss. In this case, extended conservation and protection measures are necessary to prevent disappearance of relief elements.