Chapter 10 – Microbial Ecology of Lascaux Cave
p. 253-260
Texte intégral
1This text constitutes the transcription of the presentation made by Claude Alabouvette.
2I do not have the in-depth knowledge of my colleagues from the Laboratoire de Recherche des Monuments Historiques (LRMH) on decorated caves and the microbiology of art works, which is why the ideas that I will propose here are those of a newcomer to the subject. I will present the results of a study that was above all carried out mostly by Fabiola Bastian, a Contract Research Worker financed by the Ministry of Culture via the DRAC of Aquitaine (Regional Department of Cultural Affairs).
3Thanks to the Lascaux Cave Scientific Committee, I have been able to get to know Professor Cesáreo Sáiz-Jiménez, who has excellent knowledge of the problems of microbiology in caves. We have had a considerable number of discussions, concerning the interpretation of our results in particular. As a specialist in the microbial ecology of soil and the rhizosphere, I must admit that I am not very familiar with the literature on the microbiology of decorated caves. But, as you know, you cannot interpret scientific results correctly without referring to the work of your colleagues.
4My presentation is organized in four parts:
– identification of the fungal and bacterial populations present in Lascaux Cave;
– analysis of the structure of the microbial communities in the zones selected for the “Microbiology-Microclimate” project and their evolution over time;
– influence of anthropogenic factors on the microbial communities found in the cave;
– propositions and possible guidelines actions.
10.1 Identification of the fungal and bacterial populations
5Our first goal was to identify the fungal and bacterial populations in the cave and compare the diversity of these populations in the different sampling zones.
6There was a very serious colonization of the cave in 2001 that was attributed to Fusarium solani. This was not an accidental introduction, because the work of Joëlle Dupont shows that there were several different phylotypes spread around the various compartments of the cave. This sudden development of Fusarium solani was, in all likelihood, caused by a change in the environmental and microclimatic conditions that enabled the spores—perhaps present since the origin of the cave— to develop in a visible mycelia form that threatened the preservation of the paintings in the cave. It was then decided to intervene as a matter of urgency and apply various treatments that stabilized and then very clearly improved the sanitary condition of the cave. However, as there were some persistent spots of colonization, quaternary ammoniumbased biocidal treatments—sometimes associated with antibiotics— continued to be applied until 2003, when we took our first samples. There were no scientific publications describing the native microflora in Lascaux Cave. It was therefore essential to identify the populations of microorganisms in the cave because it was impossible to imagine that only the three or four species of fungi and bacteria the most often mentioned were present in the cave.
7We finally took a series of samples in all of the cave’s compartments in April 2006; we even had the good fortune to go as far as the Shaft of the Dead Man and into the sanded-up galleries, which have never before been visited or treated and are considered to be the zones in the cave least modified by human actions. We took several samples from each compartment, making a distinction between the visibly colonized zones and the apparently non-colonized zones. One of our goals was to demonstrate that the apparently non-colonized zones are home to the same microflora as the visibly colonized zones, this question having been debated by the Scientific Committee.
8In order to identify the microflora, we used two types of methods: one consisted of culturing the organisms present in the substrate taken from the surface of the walls; the other consisted of extracting the DNA (deoxyribonucleic acid) from those substrates, without passing through the culturing stage. Starting from the cultures we identified the fungi using the conventional phenotype criteria (color of the colony, presence of the conidia, shape and size of the conidia, etc.) and also by extraction of the DNA, followed by cloning and sequencing. The bacteria were identified without passing through the culturing stage, using the DNA extracted from the supports. Whether bacteria or fungi were concerned, molecular identification consisted, after extraction of the DNA and purification, of amplifying the region of interest and then of introducing the amplicons in a plasmid cloning vector, itself transferred to Escherichia coli. The zone of interest was sequenced and a phylogenetic identification was made by comparing the sequence obtained with those deposited in the databanks. For the bacteria we used the 16S sequence of the ribosomal DNA, for the fungi the 18S sequence and also the ITS region of the ribosomal DNA. There is no reason to oppose one type of identification method to the other as they are complementary.
9The fungi that we isolated and determined according to morphological criteria were essentially Fusarium, Gliocladium, Verticillium, Aspergillus and Penicillium. As for the dominant melanized fungus in the right-hand compartment, it consisted of several species belonging to the genus Scolecobasidium. This determination was confirmed on the basis of morphological and molecular criteria by our colleague Alena Nováková, who specializes in this type of fungus. We noted that Fusarium solani, which was responsible for the intense colonizations in 2001, was only very slightly represented. It was no longer dominant within the fungal communities. On the basis of the molecular criteria, we identified a much higher number of species. Table iii presents the 10 species of fungi the most widely represented amongst the 600 or so clones that were analyzed. There was therefore a great diversity of fungi present in Lascaux Cave; the visible colonies fortunately only represent a small part of this diversity. Table iv presents the 10 most greatly represented bacterial species. These include a great abundance of Ralstonia and Pseudomonas, but also Legionella, bacteria that are often associated with amoeba and may be pathogenic for humans. The virtual absence of the species Pseudomonas fluorescens, described previously as being associated with Fusarium solani, can be explained by a poor initial identification, because the technique that was used then could not distinguish Pseudomonas fluorescens from the other species of Pseudomonas, whereas the molecular method enables this.

TABL. III – Main species of fungus: the ten most common clones per samples, by number and percentage.

TABL. IV – The most common bacterial taxa in Lascaux Cave.
10.2 Analysis of the structure of the microbial communities
10Whether fungi or bacteria are concerned, we reveal the same species from the visibly colonized zones and the apparently non-colonized zones. These results strengthen the hypothesis whereby the same microorganisms are present on the supports, and that it is highly localized environmental factors that favor the development of colonies apparent at certain points and not at other points nearby. Because fungi require a source of organic carbon, on the one hand, and favorable climatic conditions, on the other, in order to grow, these results have led to the launching of a “Microbiology-Microclimate” program whose goal is to attempt to link the microbial development observed on the surface of the walls with the microclimatic conditions also measured on the surface of those same walls. This project calls on the skills of Jean-Didier Mertz and David Giovannacci (LRMH) for the characterization of the rock’s hydric condition, of Adriana Bernardi (CNR in Padua, Italy) for measuring the microclimatic parameters at the rock/atmosphere interface, and of Fabiola Bastian (INRA), Geneviève Orial and Faisl Bousta (LRMH) for monitoring the microbial colonization.
11Three measurement, observation and sampling zones were chosen at the entrance to the Axial Gallery, because at the time we made our choice, this zone was microbiologically very active. Zones 1 and 2 are adjacent to each other on the right-hand side of the Axial Gallery; zone 1 always appears to be very dry and non-colonized, zone 2 is much damper and partially colonized. As for zone 3, it is situated to the left of the Axial Gallery, opposite the previous two zones; it is characterized by a high level of humidity and fungal colonization. Measurement instruments have been specially manufactured and placed in this spot to characterize the transfers of water in the rock, and quantify the humidity, temperature and convection currents on its surface. At the same time, samples were taken when the system was put in place and then eighteen months later to characterize the microbial populations. The measurements show that the rock, at that spot, only has very small diameter pores which are therefore always filled with water. The rock does not play an active role in the transport of water. The microclimatic conditions on the surface of the rock depend more on the climatic parameters of the atmosphere than on those of the rock. Adriana Bernardi has recorded microclimatic differences between the three study zones with, over a time, a stabilization and convergence of the parameters measured for Zones 1 and 2, which remain different from those of Zone 3.
12From the microbiological viewpoint, we have implemented all the previously described techniques for identifying the microorganisms that are present. The results show that certain fungi that are well represented in Zones 2 and 3 are absent from zone 1, the driest zone. The bacterial populations are also much more complex and diversified than in zones 2 and 3, which are the dampest. The difference affecting the diversity of the fungal and bacterial populations may therefore be correlated with the differences affecting the microclimate. We also used another methodology, called T-RFLP, which makes it possible to assess the structure of the microbial communities without having to identify the species. The results are presented synthetically on PCA planes (Principal Component Analysis). Overall, when the elements that represent the diversity of the structure of the communities appear visually superposed on the same PCA plane, it shows that the structure of those communities is not significantly different, and this must then be confirmed by statistical processing. It can be seen that, on this PCA plane (fig. 151), the points corresponding to the samples taken in February 2007 in Zones 2 and 3 are superposed but differ clearly from the points corresponding to the samples taken in Zone 1. This allows us to conclude that the structure of the microbial communities in Zone 1 is different from the structures in Zones 2 and 3, which are similar to each other.

FIG. 151 – Analysis of the main components of the structure of fungal communities in zones 1, 2 and 3: a sample taken in February 2007; b sample taken in July 2008.
13Following the meeting of the scientific committee held in the spring 2008—Adriana Bernardi having pointed out to us that the microclimatic conditions in Zone 2 were tending to converge with those of Zone 1—we took further samples to analyze the structure of the microbial communities and detect any variation that could be correlated with the change in the microclimatic conditions. The results show a change in the fungal communities between 2007 and 2008 and a modification in the structure of the communities in Zone 2, which were converging on the structure of the communities in Zone 1.
14These results therefore seem to support the hypothesis that the change in the microbial communities is determined by the change in microclimate parameters. However, it has not been possible to demonstrate this tendency rigorously. Indeed the amount of substrate available on the surface of the rock was too limited to make it possible to repeat the analyses. To perform a statistical analysis of the results, we had taken three samples per zone, but we did not have enough DNA in each sample to carry out biological repetitions, given that for each analysis we realize three methodological repetitions.
15And I would like to insist on this point, which is specific to the work in decorated caves. Our investigations are limited by the amount of material available for microbial analyses. Furthermore, each time we take samples, we destroy the object of our study, because the sample that we take six months later will obviously not be identical to the first one. And even if we only need very small amounts of DNA we are confronted with this difficulty.
10.3 Influence of anthropization on the microbial communities
16Lastly, I am going to talk about the effects of the anthropization of the cave on the bacterial communities. It is on this point that we have greatly benefited from the help of Cesáreo Sáiz-Jiménez.
17The bacterial populations in Lascaux are significantly different from those identified in non-anthropized caves. Proteobacteria represent 98% of the phylotypes identified in Lascaux, compared with 45% on average in caves that have not been anthropized. The Ralstonia and Pseudomonas genera alone represent 54% of the clones. Ralstonia and Pseudomonas are, all the same, present in the least anthropized zone of the cave, the sanded-up galleries in the Shaft of the Dead Man. This can easily be explained: human activities, particularly the treatments, do not introduce any new populations, but exercise selection pressure on the pre-existing populations and simply allow certain germs to multiply more than the others. The bacterial populations in Lascaux include a large number of species associated with protozoa, species that are representative of highly anthropized milieus such as the cooling towers of large buildings or hospital corridors.
18Immediately after the cave was re-opened, following the last biocidal treatment in January 2008, we took a sample of the earth above the ledge at the bottom of the Passageway and realized cultures in the medium enriched with 1 and 5 ml per liter of Dévor Mousse—which is the mixture of quaternary ammonium and isothiazolinone used for these treatments--and we observed growth, even in the strongest concentration, of two Pseudomonas which had been determined as being Pseudomonas brenneri and Pseudomonas chlororaphis. This result is not surprising because the literature indicates that certain species of bacteria are capable of degrading quaternary ammonium. You must be aware of the indirect consequences of these biocidal treatments used to halt the fungal colonizations. The impact of the treatment was confirmed by the calculation of the Shannon index, which makes it possible to compare the diversity of different populations. It was much higher in the colonized and therefore frequently treated zones, than in the non-colonized zones. The treatment contributed to increasing the diversity of the bacterial populations, favoring the development of certain species that are not greatly represented within the populations in the non-treated zones.
19Concerning the fungi, the Fusarium solani isolated in the cave were resistant to formaldehyde, which is quite logical given the large number of treatments applied with that product in the past. It must be pointed out that Fusarium develops from any source of organic carbon including ethanol. The colonization by Scolecobasidium, very clearly dominant in the right-hand Passage (we have identified it from 14 different samples) could also be associated with the frequency of biocidal treatments because this fungus is described in Japanese publications as being capable of developing using the detergents in laundries, for example, as their sole source of carbon and nitrogen.
10.4 Possible guidelines
20Given this inventory, what indications do we have to ensure the preservation of the art works in Lascaux Cave? Yesterday, we talked a lot about climatic equilibrium, and today I believe we must talk about microbiological equilibrium. Soil microbiologists know that all these bacterial and fungal populations interact permanently in the soil, and pathologists also know that pathogens are never eradicated. It is therefore necessary to act on microbial balances to prevent the pathogen from developing and its harmful activity from appearing. This is, I think, what we should try to do in Lascaux. We must manage to establish an equilibrium between the microbial populations such that it prevents some of them from proliferating. It is not a question of restoring the original microbial equilibrium, which is lost and will no doubt never be restored, but an equilibrium that must make it possible to stop or even annihilate the development of the fungi that are dangerous for the preservation of the paintings.
21As in agriculture, chemical control has many drawbacks. We know that the microorganisms form biofilms that combine several species of bacteria and fungi on the surface of the cave’s walls. The properties of these biofilms are different from those of each species cultivated in a pure culture medium. It has been demonstrated that the same species are more resistant to biocidal treatment when they are associated in the biofilms than when they are in a pure culture medium, and we must question the validity of the laboratory tests used to evaluate the effectiveness of the different biocides.
22The bacterial populations in Lascaux include several species that we know are associated with protozoa and fungi. In particular we have found a bacterium called Collimonas, which is mycophagous, that is to say capable of developing at the expense of fungi; furthermore it has been proposed to use it for biological control purposes. But this bacterium is also well known to researchers who are interested in the degradation of minerals, because it is autochemolithotrophic, that is to say capable of developing by drawing its energy from minerals. It may therefore be at the origin of the biological cycle, because it is capable of developing very slowly on minerals and at their expense. The dead bacterial cells then constitute organic matter that can be used by another microorganism. This example shows the complexity of the interactions between microorganisms and between microorganisms and their surroundings. We must therefore intervene with the greatest caution as any action could have unsuspected consequences.
23It is not possible to extract the organisms that pose a problem, that is to say Fusarium solani and Scolecobasidium, from their biological context. We must accept the difficulty of taking account of all the microbial communities that depend on the microclimate, and that is where the whole difficulty lies. Amongst the fungi identified in Lascaux there are species known for being transported by arthropoda—which can ingest spores that, once they have been expelled, germinate more easily in a Petri dish—and other species liable to destroy those same arthropoda and that are used in biological control. There again, the interaction is one-way and any sudden intervention could have unforeseen consequences. Lastly, the residues of the biocidal products could represent a source of carbon and nitrogen that could be used for the surviving microorganisms. The results of the courses of treatment are absolutely convincing and once the treatment has been applied, 95% to 99% of the microflora are destroyed. Nevertheless, what is left is only asking for a chance to start up again. For example, until recently, farmers were allowed to treat the soil with an extremely dangerous molecule (methyl bromide) to disinfect it. However, they then had to disinfect it again every year because the pathogen reappeared each year. We must therefore give up any desire to eradicate the elements that bother us; we must simply try to find a way to prevent them from developing and being harmful.
24Yesterday, we wondered about the origin of the melanin on the walls. Is it the same as that produced by the black fungi that are posing a problem? What is this melanin’s lifespan? What is the physiology of these fungi? What are the conditions under which these fungi produce the melanin? Can physical means be used, such as germicide radiation, to destroy the fungi without damaging the works? These are all questions that are worth consideration. We must also pose the question again of the vectors (insects, humans and convection currents) that can transport the fungal propagules, because the air-contamination monitoring performed by LRMH has never made it possible to detect in the air the fungi that are present on the walls.
25Lastly, we come to the really fundamental question: how to establish or re-establish a microbiological equilibrium that would be favorable for the conservation of art works? In my opinion, this will only be possible by adopting a holistic approach and better climate regulation. Eventually, the microclimate measurements concerning a few square decimeters of wall surface will have to be linked to the model of the prototype managed by Delphine Lacanette (cf. chapter 6). I think that is the direction we must follow.
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Mottes castrales en Provence
Les origines de la fortification privée au Moyen Âge
Daniel Mouton
2008
Géoarchéologie de sites préhistoriques
Le Gardon (Ain), Montou (Pyrénées-Orientales) et Saint-Alban (Isère)
Dominique Sordoillet
2009
L’enceinte des premier et second âges du Fer de La Fosse Touzé (Courseulles-sur Mer, Calvados)
Entre résidence aristocratique et place de collecte monumentale
Ivan Jahier (dir.)
2011
Lyon, Saint-Georges
Archéologie, environnement et histoire d’un espace fluvial en bord de Saône
Grégoire Ayala (dir.)
2012
Les gisements précolombiens de la Baie Orientale
Campements du Mésoindien et du Néoindien sur l’île de Saint-Martin (Petites Antilles)
Dominique Bonnissent (dir.)
2013
L’Îlot du palais de justice d’Épinal (Vosges)
Formation et développement d’un espace urbain au Moyen Âge et à l’époque moderne
Yves Henigfeld et Philippe Kuchler (dir.)
2014
Bettencourt-Saint-Ouen (Somme)
Cinq occupations paléolithiques au début de la dernière glaciation
Jean-Luc Locht (dir.)
2002
Campements mésolithiques en Bresse jurassienne
Choisey et Ruffey-sur-Seille
Frédéric Séara, Sylvain Rotillon et Christophe Cupillard (dir.)
2002
Productions agricoles, stockage et finage en Montagne Noire médiévale
Le grenier castral de Durfort (Tarn)
Marie-Pierre Ruas
2002