Chapter 9 – Managing Biological Activities in Lascaux: Identification of Microorganisms, Monitoring and Treatments
p. 220-251
Texte intégral
9.1 Introduction
1The preservation of paintings and engravings in natural sites such as decorated caves is probably the most difficult challenge facing those responsible for heritage conservation. In this domain, France has a very important legacy and Lascaux Cave, a UNESCO World Heritage Site, is one of the most remarkable examples. It is located in the district of Montignac (Périgord) and is one of the most famous prehistoric painting sites in the world. However, since its discovery in 1940, it has gone through several periods of crisis, which, among other things, have disrupted its biological equilibrium. The first problems of a biological nature date back to the 1960’s, with the abnormal development of green algae cells on the supports, which led to the cavity being closed in 1963. In September 2001, the cavity suffered an unprecedented contamination of molds: in just a few days a white blanket of mycelium covered all the floors, lateral ledges and part of the sloping surfaces, requiring emergency treatment. In July 2007, another episode of contamination, in the form of a black colonization, developed more particularly in the right‑hand part of the cave, requiring the implementation of another biocidal treatment.
2The LRMH (Laboratoire de Recherche des Monuments Historiques) has been monitoring the microbiological changes in Lascaux Cave for more than thirty years by means of regular verifications of the conditions of the atmosphere and walls and, when necessary, it makes treatment recommendations.
9.2 The preservation of rock art, a complex problem
3All natural cavities, whether decorated or not, represent open systems in constant contact with their internal and external surroundings. The environmental conditions, around 12 °C and 99% relative humidity, are in theory severe for the preservation of paintings; the very high humidity, close to the saturation point, combined with periods of condensation, promotes trickling, causing the pigments to run. Likewise, the natural presence of carbon dioxide is not without risks for the preservation of the rock paintings. The sometimes high levels of CO2 may lead to the dissolution of the carbonate substrates due to the formation of carbonic acid, or the precipitation of calcite on the walls with the formation of translucent or opaque veils (Brunet, Vouvé 1996).
4Although relatively stable environmental conditions have made it possible to preserve certain works until the present day, there is little humans can do to the cave environment to “improve” the preservation conditions of this art: any significant lowering of the relative humidity can cause certain pigments to dry out and turn powdery and any significant inflow of water, if not counterbalanced by temperature control, can lead to condensation and trickling.
5As regards biology and microbiology, these cavities have specific flora and fauna that may vary in time and space, depending on the endogenous natural inflows (water infiltrations, etc.) or exogenous inflows linked to human activities. They create situations that are extremely favorable for the development of microorganisms:
– water‑rich substrates such as clays and/or substrates with highly undulating surfaces (potential niches for microorganisms),
– presence of organic matter.
6The developments observed (algae, fungi, lichens) usually remain limited due to the biological balance that has established itself over the long term. Sometimes, due to climatic or environmental events, one biological species may develop anarchically, and once that development has begun, it is very difficult to stop it because of the highly favorable initial conditions (inaccessible niches, reseeding; Roquebert, Orial 2002).
9.3 Monitoring the condition of Lascaux Cave until 2000
9.3.1 Characteristics of the cave
7Lascaux is a descending type cave whose entrance is situated in the upper part of the hill. It is a medium‑sized cavity, approximately 60 m in length from the entrance to the end of the Axial Gallery and approximately 110 m from the entrance to the bottom of the righthand Gallery. The accessible system has a volume of about 320 m3 for the Hall of the Bulls and 1,300 m3 for the right‑hand part (Apse and Nave).
8The decorated walls are about ten meters from the ground outside. They cover 54 m2 in the Hall of the Bulls, 100 m2 in the Axial Gallery and 120 m2 in the Nave and Apse (fig. 105). At present, the temperature of the air inside the cavity is of the order of 12.5 °C and the relative humidity is close to the saturation point.

FIG. 105 – Left wall of the Hall of the Bulls.
© D. Bouchardon / LRMH.
9.3.2 The first biological crisis in the 1960’s
9The first spot of green algae appeared in September 1962 after twelve years of uninterrupted visits, causing a surplus of CO2 and light. Faced with the reappearance of these green growths and the appearance of veils of white calcite, the cave was closed to the public in 1963 due to what was called the “green disease” and the “white disease”. The agent responsible for the contamination was identified by Doctors Lefevre and Pochon from the Institut Pasteur (Lefèvre, Laporte 1969; Lefèvre 1974), as being a single unicellular alga, a Chlorophyta and more precisely of the genus Bracteacoccus (fig. 106).

FIG. 106 – Bracteacoccus sp.
© LRMH.
10However, analysis of the sediments and walls revealed the existence of a multitude of living plant and animal organisms—bacteria, algae, and molds, as well as small colorless flagellates, amoeba, etc. The first conclusions drawn by scientists were that the cave had certain selective properties: although abundantly present, the bacteria, fungi and algae do not develop and remain at the spore stage. Only one species of algae was capable of multiplying there. The notion of self‑purification was even proposed (Lefèvre 1974).
11The algae decontamination process was then defined, involving the application of a colorless biocide that would have no impact on the paintings. A 1:200 solution of formaldehyde was chosen and applied to the bare rock, the floor and the paintings. In four months, the green algae were eliminated from the cave. A preventive treatment was then recommended and implemented. It consisted of localized spraying of formaldehyde on the paths and the installation of a footbath, also filled with formaldehyde to disinfect the shoes of visitors, at the entrance to the cavity. These interventions continued until 1998. It should be noted that until the present day there has not been any reappearance of green algae.
129.3.3 Microbiological monitoring until 2000
13Until the 1970’s, monitoring of the microbiological pollution was carried out by the Institut Pasteur. The task of sanitary monitoring was then passed to the LRMH. This consists of evaluating the microbiological pollution (bacterial, fungal and algal) of the air, associated with a visual observation of the condition of the accessible surfaces.
14For about twenty years, a “static” method was used. Petri dishes containing culture media specific to each type of microorganism were placed on the floor and left open for 5 minutes for the bacteria and fungi, and for 24 hours for the algae. During this period, the microorganisms present in the air settled on the culture medium under the effect of gravity. At the end of the exposure time, the dishes were closed and taken back to the laboratory for incubation in optimum culture conditions. The viable and cultivable organisms were counted and expressed as Colony‑Forming Units (CFU). Some ten points chosen inside the cavity and in the entrance airlock chamber were checked using this protocol twice a year, until 1990. The rates of organisms present in the air varied at that time within very low ranges, from 0 to 5 CFU for molds, and from 0 to 15 CFU for bacteria. The results showed that the algae and bacteria pollution decreased spontaneously if the visits were suspended, a phenomenon indicative of the cave’s selfpurification capability.
15The method was modified in 1990 when the so‑called “dynamic” mode was introduced. The culture media remained the same as for the static method, but the Petri dishes filled with agar culture media were placed, open, in a sort of aspirator. During the suction time, the microorganisms transported in the aspirated air came into contact with the agar. The suction operation, which lasted a few seconds, was repeated for each type of microorganism with the appropriate culture medium. The dishes were then closed and taken back to the laboratory for incubation, followed by colony counting. As the volume of aspirated air is known, the CFU number can be calculated per cubic meter of air (figs 107‑109).

FIG. 107 – Sampling of air in the Axial Gallery.
© LRMH.

FIG. 108 – SAS Compact equipment.
© LRMH.

FIG. 109 – Example of a culture medium after the sampling of air and incubation.
© LRMH.
16With dynamic monitoring, the basic level is higher but remains within low ranges, between 0 and 50 CFU/m3 for molds, and between 0 and 150 CFU/m3 for bacteria. This dynamic method is still used today and, to allow for the reproducibility of the results, the protocol remains the same for each measurement session: apparatus, culture medium, suction time and location of the sampling points.
17The pollution detected in the air is mainly linked to human movement and activities. The pollution that may be observed on the supports is generally more closely associated with a development of telluric organisms. It must be highlighted that the accumulation of aerial pollutants intensifies the surface biofilm, as they settle, thus creating niches favorable to the emergence of other telluric microorganisms.
18Some minor imbalances are sometimes observed—for example at the time the replica was made in 1973. However, thanks to the limitation placed on the number of visitors (5 per day) and the diffuse lighting, the equilibrium of the ecosystem was maintained and, until 1994, no visible microbial manifestation was detected in the cavity. At that moment, the first small deposit of fungal origin to appear in the cavity (on the floor in the Hall of the Bulls) was recorded by the monitoring technicians. It was also at that time that the aerial pollution began to fluctuate, sometimes with sharp peaks, but also with regular returns to the base levels. We must keep in mind that there is no benchmark available for this type of monitoring. Each decorated cave—depending on its size, volume, exchanges with the surroundings—will have its own “microbial identity card”, a sort of natural biological “background noise” and also a capacity to “accept” or “withstand” a certain degree of exogenous pollution. We can call this self‑purification or regulation with a limit threshold, below which the equilibrium is maintained and above which there may be disruption in the form of visible biological manifestations. The problem lies in our lack of knowledge of the theoretically acceptable limits, which are specific to each cavity.
19For more than thirty years, the cavity did not show any worrying signs of disruption and, despite the sporadic appearance of biological deposits and the variability of the aerial biocontamination linked to greater frequentation of the cavity, there was nothing to suggest the sudden, massive fungal contamination that occurred at the end of 2001.
9.4 The contamination in 2001 and how it was managed
20In September 2001, after the renewal of the air conditioning system, the cavity fell victim to unprecedented mold contamination: a white blanket of mycelium covered all the walls, the lateral ledges and part of the sloping surfaces. This blanket developed first in ‘Airlock 2’, then in the Hall of the Bulls and the Axial Gallery and lastly throughout the Nave (figs 110‑112).

FIG. 110 – Development of white mycelium on a bench.
© LRMH.

FIG. 111 – Detail of white colonies on a rock scale.
© LRMH.

FIG. 112 – Appearance of colonies on clayey sediment.
© LRMH.
9.4.1 Identification
21The strain responsible for the contamination was identified as being Fusarium solani, a telluric wet‑spore species enveloped in hydrophilic mucus and transported by water. This species, whose activity level in water (a w) is high (> 0.9), is very common in soil and is known in the agricultural sector as being responsible for the plant disease called “fusarium wilt” (fig. 113).

FIG. 113 – Microscopic view of Fusarium solani (M x 400).
© LRMH.
22Examinations with stereoscopic microscopy and scanning electron microscopy of small flakes of rock collected on the site and colonized by the Fusarium show the intensity of the mycelial cover and the depth of penetration of the hypha in the substrate (fig. 114, 115). These observations also give a clearer understanding of the physical damage that may be caused by molds on the colonized supports, not forgetting the chemical alterations linked to metabolic secretions such as organic acids.

FIG. 114 – Appearance under magnification of the colonized surface of a rock scale.
© LRMH.

FIG. 115 – Scanning electron microscope view of the same scale: visualization of mycelian filament penetration of the substrate.
© LRMH.
23Only an immediate intervention could avoid the propagation of the colonies to the paintings themselves, with the obvious risks for their conservation.
9.4.2 Initial treatments
24The LRMH therefore proposed an emergency treatment that consisted of spraying a fungicidal product based on quaternary ammonium, known for its fungicidal properties and used for many years in the heritage sector. This chemical treatment was associated with the physical removal of mycelia (Orial, Mertz 2006a).
25This first treatment was unfortunately not effective: the surfaces in the cave are very uneven and difficult to access and the mycelia developed more quickly than the treatment could cope with. A substitute treatment had to be found very quickly.
26After having evaluated the potential risks, the LRMH proposed to apply quicklime to the floors in order to block the progression of the mold. There were several reasons for this choice:
– the disinfectant properties of quicklime, which have long been known, linked to the increase in the pH of the substrate and the absorption of the water in which the microorganisms developed during lime slaking;
– the good chemical and mineralogical match with the karst substrates;
– the possibility of carrying out a general and fast mass treatment to stop the uncontrolled extension of the mycelia.
27This application was carried out at the end of October 2001, and a check of the hydration and carbonation of the lime was realized (Orial, Mertz 2006a). The contamination of the floors was thus contained through the combined action of the dehydration of the microorganisms and of the brutal rise in the pH when the lime was slaked. This caused the macroscopic drying of the mycelial filaments and, as a corollary, the drainage of the earth floors by reducing their water content. The molds were effectively inhibited on the floor, but it was not possible to treat all the zones in this way, and there remained spots of active development.
28An effective fungicide had to be selected that was not toxic for the paintings and that could be applied by spraying all the vertical zones or those areas that were difficult to access. A large number of fungicidal products were then tested in the laboratory using the anti‑biogram method (fig. 116), on an agar medium and in microplates. The effectiveness of some twenty formulations was evaluated on the Fusarium strain, such as the quaternary ammoniums with different carbon chains, ortho‑phenylphenol, glutaraldehyde, carbendazim, isothiazolinones and imidazoles. Formol, which has been used for many years as a preventive treatment, was also tested without any convincing results. The best results were obtained with the quaternary ammonium based on alkyl dimethyl benzyl ammonium chloride recommended right from the beginning of the crisis and made up of an alkyl chain C12 and C14.

FIG. 116 –Examples of antibiograms.
© LRMH.
29In parallel, all the other treatment hypotheses were explored: from gassing with methyl bromide or inhibition by raising the level of carbon dioxide, or the use of biological control measures. The choice of nonintervention was also considered in the hope of an eventual return to the natural equilibrium. The advantages of each of the solutions were analyzed but none of them could objectively be adopted in view of the excessive risks inherent to each one. The solution of chemical treatment with quaternary ammonium was therefore the only one that could be used without risk, despite the application difficulties; this decision was confirmed by in‑depth bibliographic research attesting to the harmlessness of these products on works of art (Petersen et al. 1993).
30It is important to insist on the difficulty of treating this type of very humid, uneven rocky surface, with numerous crevices: fungicide solutions cannot penetrate into a porous network already saturated with humidity, and not all the contaminated zones were accessible. This is what justified the development of various application techniques such as compresses (fig. 117), or a trial treatment with quaternary ammonium paste.

FIG. 117– Compresses applied to a sloped surface in the Hall of the Bulls.
© LRMH.
31Despite all these interventions, the cavity’s biotope was still disrupted, which was materialized, for instance, by local appearances of bacterial mucus linked to the development of Pseudomonas fluorescens and which required a specific anti‑bacterial treatment (streptomycin + polymyxin), applied temporarily for several weeks. Likewise, small round colonies mainly due to the development of a pigmented mold, Gliomastix murorum, appeared from time to time, mainly on the ceilings (Airlock 2, Apse, Hall of the Bulls: fig. 118). These zones received the same fungicidal treatment as the walls. To date, these stains are still visible but, overall, they are no longer developing.

FIG. 118 – Appearance of black colonies on the ceiling of the Hall of the Bulls.
© LRMH.
32Starting from January 2004, and given the significant regression of the cavity’s overall contamination, the scientific committee decided to modify the treatment protocol and replace the chemical treatments with cleaning and mechanical elimination operations involving, in particular, the development and use of a vacuum injection‑extraction system for the undecorated and non‑fragile zones, the priority goal being the physical removal of the filaments and, as a consequence, of the reservoirs of contaminating spores transported by these mycelia. Compresses or sponges were also used widely on the most fragile substrates (fig. 119, 120).

FIG. 119 – Removal of the black colonies using a sponge.
© A. Moskalik-Detalle.

FIG. 120 – View of the mycelia removed. © LRMH.
9.4.3 Scientific monitoring of the condition of the cave
33The cave’s condition was monitored using a wide range of analyses:
– regular macroscopic observations of the surface condition;
– nearly 200 samples of mycelium or suspicious deposits taken to identify the species present;
– more than 60 samples of soil, clays and sediments taken for qualitative and quantitative microbiological analysis;
– some twenty atmospheric pollution verifications throughout the cavity (15 points systematically checked).
9.4.3.1 Analysis of surface swabs
34From the beginning of the crisis and for the following years, samples were taken—regularly and each time they appeared—of the suspicious deposits observed on the floors, walls and sometimes on the paintings by means of sterile swabs in order to identify the species responsible for the visible colonies.
35The mycelium samples were cultured in the laboratory on synthetic media specific to molds (malt agar) for their morphological identification. The results prove that it is Fusarium sp. that is most often responsible for the visible white contamination, and also Verticilium sp., Penicillium sp., Aspergillus sp., as well as Gliomastix sp., Gliocladium sp., Cladosporium sp. and Alternaria sp. for the fungi with black melanin pigmentation (fig. 121, 122).

FIG. 121 – Verticillium sp. (M x 400).
© LRMH.

FIG. 122– Gliocladium sp. (M x 400).
© LRMH.
9.4.3.2 Analysis of sediments Methods
36The quantitative and qualitative evaluation of the microorganisms present on or in a rocky medium can be carried out by means of conventional counting and identification methods (Roquebert, Orial 2002). As far as the bacterial organisms are concerned and contrary to medical bacteriology, which is based on the examination of isolated strains, the evaluation of the floors or rocks is above all based on the examination of a heterogeneous population, with important physiological functions making it possible to classify the bacteria in functional groups. Within these groups, the reactions are organized in complex and overlapping series (oxidation, reduction, mineralization, synthesis, etc.) that form chains or more or less closed cycles. In general, there are too many microorganisms in a sample to be quantified using the basic sample. The preparation of a suspension‑dilution of the substrate for analysis in sterile water is then required and is used to seed different culture media, favoring the selection of the functional groups. This stimulated culture makes it possible to count the viable organisms and establish a sort of species viability statistic with the definition of the most probable number of organisms present. The figures obtained are then compared with the Mac Grady statistical tables, which make it possible to evaluate the most probable number of organisms per gram of sample.
37The identification of the bacterial species will provide a useful complement to the counting. This identification is based on the comparison of the various phenotypic characters (morphological, biochemical, physiological) of the strains to be studied with baseline strains. The automation of these identification methods, using galleries such as the API system, has the advantage of standardizing the biochemical characters being sought and then of improving the reproducibility of the results. Nevertheless, these techniques are only suited to the diagnosis of cultivable organisms because the prerequisite for identification is that you must have a pure culture of the species. On the soils and sediments of Lascaux, it is mainly the total heterotrophic microflora, algae and microscopic fungi that are quantified and identified. For some samples, the organisms entering into the nitrogen, sulphur, iron and manganese functional groups are analyzed.
38The insoluble nitrogen and the total nitrogen are measured using the Kjeldahl procedures, from which the soluble nitrogen content is deduced. The nitrogen compounds in the dry powder samples are mineralized by means of concentrated sulphuric acid, containing a high concentration of potassium – which makes it possible to raise the mixture’s boiling point – and in the presence of selenium (catalyst), to form ammonium sulfate. The nitrogen is then released in the form of ammoniac thanks to the addition of concentrated soda, then distilled in a solution of boric acid indicator. Titration is used to determine the amount of ammoniac, by means of a burette, with sulphuric acid that turns the green boric buffer pink, in excess (back titration). The results are expressed in milligrams of nitrogen per gram of dry matter. Three repetitions are carried out for each sample. Unfortunately, the amount of matter taken in Lascaux did not make it possible to perform the three repetitions.
39The organic carbon content (OCC) is determined by calculating the difference between the total carbon and the mineral carbon. Analysis is carried out with an analyzer (Shimadzu TOCmeter) using dry powder that is carbonized at 200 °C. The apparatus gives the weight of the total organic carbon directly expressed as a percentage of the sample’s dry weight. Three repetitions are performed for each sample.
40Results
41For all sixty samples of soil and sediment taken throughout the cavity at the beginning of the crisis and over the following years, the total microflora is within a range that is quite normal for soils, of 104 to 108 CFU/gram of matter. Bacterial diversity is not particularly rich. The most frequently found species are: Pseudomonas fluorescens, Agrobacterium radiobacter, Stenotrophomonas maltophilia, Ochrobacter anthropi and Brevibacillus brevis, which are common telluric species in all soils. The mold spores are present in average quantities of between 0 and 106 CFU/ gram of soil with not much diversification in the genera: Fusarium sp., Penicillium sp., Aspergillus sp., Cladosporium sp., Gliomastix sp. and Trichoderma sp. Algae were not isolated in any of the samples.
42Figure 123 shows an example of a CFU/gram count of soil in a solid medium of the total microflora (bacteria, fungi) and of the algae in three samples. Two solid media – soil extract and Czapeck‑Dox – were used to count the total microflora while two other media (nutritive agar and malt‑agar) were used to count the bacteria and fungi respectively. Certain organisms intervening in the functional groups of the nitrogen and sulphur have been quantified. For example, those used in the ammonization and oxidation of the nitrogen (nitrozation and nitration) can achieve maximum values of 108 organisms/gram of sediment (histogram showing the results of certain analyses). These organisms help to enrich the soil with nitrogen compounds and correspond to mineralization. The denitrifying agents, on the other hand, reduce these compounds into nitrogen. It so happens that the latter are in the same value range as the mineralizing agents. There will therefore not be any enrichment or impoverishment of the soil in nitrogen. Furthermore, the measurement of the total organic nitrogen gives values of around 0.1‑0.2 g/l and the measurement of the C/N ratio gives low values, not exceeding 17. As far as the various sulphur element transformation processes are concerned, the evaluation of the mineralizing agents was negative, and the sulphur oxidants (thiobacillus) did not exceed 104 organisms/gram of sediment. There were no organisms contributing to the oxidation of iron or manganese. Figure 124 shows an example of the analysis of the microflora entering into the functional groups of nitrogen, sulphur, manganese and iron for the previous three samples.

FIG. 123 – Counts of the bacteria, fungus and algae.
© LRMH.

FIG. 124 – Germ count for germs in the operational groups of N, S, Fe and Mn.
© LRMH.
43The determination of the sediments’ microbial spectrum makes it possible to conclude that the soil in the cave has a low to normal quantity of telluric flora and the carbon/nitrogen ratio shows that the soil is poor.
9.4.3.3 Monitoring of the biocontamination of the air
44Since 2001, the number of points monitored in the cavity has been more than doubled from 7 to 15 to make it possible to better assess the level of pollution of the air. The exact location of the sampling points is shown in fig. 125.

FIG. 125 – Plan of the cave with the location of the air contamination checkpoints: the points checked over thirty years are indicated in red, and those added at the time of the 2001 crisis are indicated in blue. Plan de la grotte :
© N. Aujoulat / ministère de la Culture et de la Communication, CNP.
45In order to better visualize the changes in the atmospheric pollution levels, a contamination scale has been established on the basis of all the results of the atmospheric pollution measurements obtained over the last thirty years, a period during which the cavity could be considered to be healthy or at least in a state of equilibrium. This scale has also been defined according to the experience we have gained in this respect from other decorated caves.
46In order to attempt to provide a more synthetic image of changing contamination levels and thus highlight trends, the measurement brackets have been color‑coded, with six levels of contamination being defined according to the following code:

47As far as mold is concerned, the contamination level illustration (fig. 126) shows that, most of the time, the air in the cavity remained at very low, or even zero levels of contamination after February 2002 and, very clearly so, for example between January and June 2003. Then, in March 2004, an upturn can be seen in the cavity’s level of contamination, with high to very high levels being reached throughout the cavity in December 2004. Since then, the contamination of the air has decreased regularly and is currently at a low to very low level without ever returning to the zero level of March 2003. The identification of the species shows that, overall, until March 2004, the few molds present in the air were not particularly diversified. Some Fusarium sp., Cladosporium sp. or Gliomastix sp. were isolated from time to time. After that, the number of different species increased significantly, predominantly with Penicillium sp. and Aspergillus sp., which were found throughout the cavity. Unlike Fusarium, these species are airborne and transported, for the most part, by the various movements involved in human activity.

FIG. 126 – Trend in the fungal pollution level from November 2001 to February 2009.
© LRMH.
48With regard to bacteria, the contamination level illustration (fig. 127) shows the period between June 2002 and January 2003 as being markedly more polluted than at the beginning of the crisis in November 2001. The level then decreased during the course of 2003 and increased again in March 2004 to reach very high levels in December 2004 (fig. 128). Since then the bacterial contamination has fallen notably and returned to low to average levels. The species isolated are predominantly Micrococcacceae, Bacillus sp. or staphylococcus, linked to human activity.

FIG. 127 – Trend in bacterial contamination from November 2001 to February 2009.
© LRMH.

FIG. 128 – Chromatogram of quaternary ammonium specimen and of the same quaternary ammonium after contact with Stenotrophomonas maltophilia.
© LRMH.
49These data confirm that the fungal contaminations observed on the floors and walls, leading to visible manifestations of mycelia, are not correlated with any pollution of the air. When the contamination of the walls with Fusarium solani was at its worst (February 2002), the air did not have any fungal pollution.
50The increase in the levels of contamination observed starting in March 2004 and during the year 2005 is directly related to a high level of human activity. This is why we must remain extremely vigilant as regards behavioral attitudes and the number of people entering the cave. It must be said, however, that the aerial pollution measurements give an image at a precise instant, and this may fluctuate significantly depending on the events that preceded the measurement. Furthermore, the cave is a closed space and, as the spores cannot leave this space, they only have one possibility: settling on the surfaces. When there is a decrease in the measured values, after an increase, this simply means that the spores have settled on the walls, thus contributing to the formation of a biofilm, and that they are potentially ready to germinate.
9.5 Additional investigations
9.5.1 Additional information about the Fusarium solani species isolated in the cavity
51From the beginning of the contamination, numerous questions were raised concerning Fusarium solani. Do all the strains belong to the same species? What diversity exists within the species? What is the distribution of the species and strains in the cave? Has there been a massive invasion of a single strain of Fusarium, introduced from the outside and which has propagated throughout the cave subsequent to the work carried out in 2000?
52Additional molecular biology investigations were conducted by Joëlle Dupont, Professor at the Museum d’Histoire Naturelle of Paris (Dupont et al. 2007). Thirty-six colonies of Fusarium were isolated from the samples taken from five locations in the cave: Hall of the Bulls, Passageway, Apse, Nave and the Mondmilch Gallery. Analysis of the genetic polymorphism suggests the presence of three species (Sp1, Sp2 and Sp3) of the Fusarium solani complex. Of the 36 isolates, there were 19 strains distributed as follows: 11 strains of species 1, 2 strains of species 2, 6 strains of species 3. The zones situated at the bottom of the cave (Mondmilch Gallery and Nave), which are less frequented, appear to be home to more homogeneous communities than those located closer to the entrance. Most of the genotypes seem to be “local”, for example the five genotypes of species 1 found in the Hall of the Bulls are different from those found in the Nave at the same moment.
53The contamination is therefore multiple and dispersed, and it is not a case of invasion by a single strain that has propagated throughout the cavity. Fusarium solani may have been introduced into the cave by several separate or combined pathways (percolation from the top soil, external soil brought in by visitors, small animals or run-off water). The most plausible hypothesis is that the predominant species of Fusarium throughout the cavity have benefited, at some precise moment, from micro-environmental conditions favorable to their germination, more particularly when the air treatment unit was being replaced. The micro-climatic changes caused the simultaneous germination of the different strains of Fusarium.
9.5.2 Study of the biodegradability of quaternary ammonium by the telluric organisms specific to the cave
54The utilization of quaternary ammonium salts for biocide treatments raised a certain number of questions within the committee concerning, in particular, the possibility of biodegradability by the cavity’s telluric organisms. Experiments were therefore conducted in the laboratory to confirm or invalidate this hypothesis. It must be remembered that in the literature there are a large number of publications which examine the effects of quaternary ammoniums on the microorganisms, and several authors have thus attributed the microbial activity of these salts to various mechanisms:
– the more or less selective denaturation of the proteins or enzymes, by solubilization and depolymerization, leading to the inactivation of the enzymes involved in the respiration and glycolysis and inactivation of the dehydrogenase;
– attachment to ribosomes, with halting of the protein synthesis;
– lysis of the cellular membrane, with disruption of the osmotic exchanges.
55The bactericidal activity of alkyl ammonium compounds and particularly of alkyl benzyl ammonium chloride (C12‑C18), was suggested for the first time by Domagk in 1935.
56Our experiments were carried out on four microorganisms isolated from the clay soils in Lascaux Cave, but only the results obtained with two types of bacteria, Stenotrophomonas maltophilia and Bacillus sp., are presented here. Three types of investigations were conducted:
– culturing on a minimal medium in which the only sources of carbon and nitrogen, essential for the development of the microorganisms, were provided by the quaternary ammonium;
– high performance liquid chromatographic (HPLC) analysis of the biocide after contact with the microorganisms to evaluate their ability to degrade the product;
– lastly, examination by transmission electron microscopy (TEM) of the impact of the quaternary ammonium on the microorganisms, by visualization of their morphology. For analytical purposes, the bacteria were placed in contact with the biocide product, in a liquid medium. Then, after filtering, the bacteria were used for culturing and the TEM observations, and the filtered solution was recovered for HPLC analysis.
9.5.2.1 Culturing on a liquid and solid minimal medium
57In a liquid medium, the possible degradation of a biocide product can be verified by observing the growth or survival of microorganisms. The results show that the microorganisms are incapable of developing in such a medium. Furthermore, when these microorganisms have come into contact with the biocide and are reseeded on media free from toxic products, they do not develop. On a solid medium, the degradation of the biocide product is shown by the appearance of a translucent halo around the colony, indicating the degradation of the product. No phenomenon of this type was observed. The results show that the microorganisms are incapable of developing on a minimal solid medium with quaternary ammoniums added.
9.5.2.2 HPLC analysis
58These analyses were carried out in cooperation with Witold Novik, a biochemist at LRMH. The chromatogram shows the disappearance of certain chains of the product after contact with the microorganisms. However, no newly grown products or products resulting from a possible partial degradation were detected (fig. 128).
59Two peaks observed in the spectrum of the quaternary ammonium in contact with a preculture of Stenotrophomonas maltophilia, corresponding to two long chains, C14 and C16, disappear completely, and the dominant peak of the C12 chain decreases very significantly.
60The hypothesis can therefore be put forward that the fatty chains of the product that have disappeared have been completely degraded. No other product that could be the result of the degradation of the biocide was detected.
9.5.2.3 Examination by transmission electron microscopy
61To understand the disappearance of the chains, the titration of the nitrogen in the walls and inside the cells (cytoplasm) of the various microorganisms was performed by means of transmission electron microscopy, coupled with electron energy loss spectroscopy on the transmitted electrons. The examinations were carried out in cooperation with Dr Jaâfar Ghanbaja, from the Joint Electron Microscopy and X Microanalysis Department at the Henri‑Poincaré Science Faculty in Nancy.
62The observations on ultra‑thin sections make it possible to compare the morphology of the various microorganisms before and after contact with the biocide. It can be seen that the cells that had come into contact with the product had undergone significant modifications to their walls and membranes, their genetic material and their size, by comparison with the control microorganisms (fig. 129‑132). Salton, in 1951, suggested that quaternary ammoniums cause the death of the cells by interaction with the cytoplasmic membranes. Later, in 1984, Isomaa tried to explain the mechanism of action of surfactants (C10‑ C20) and proposed that the target on which the surfactants act is the cells’ double phospholipid layer. The figures show clearly that the bacteria’s membranes were degraded and that the cells were emptied of their content, which explains the decrease in their size.

FIG. 129 – Stenotrophomonas maltophilia (M x 20 000).
© LRMH.

FIG. 130 – Stenotrophomonas maltophilia after contact with the Vitalub QC 50 (M x 20 000).
© LRMH.

FIG. 131 – Bacillus sp. specimen (M x 20 000).
© LRMH.

FIG. 132 – Bacillus sp. after contact with the Vitalub QC 50 (M x 20 000).
© LRMH.
63The calculation of the cytoplasm’s N/C ratio and of the bacteria’s membrane structures in contact with the quaternary ammonium shows a slight increase in the nitrogen rate (2%). This slight increase can be attributed to the low concentration of quaternary ammonium used, of the order of 0.5 ppm of active matter; it is essentially localized at the cytoplasmic level.
64This study shows that the quaternary ammoniums used are not biodegradable by the microorganisms in Lascaux Cave. The disappearance of certain biocide chains, evidenced by the HPLC technique, is explained by the preferential adsorption of those chains on the membranes and walls of the microorganisms. The adsorption of the biocide molecules on the membrane structures of the microorganisms causes a decrease in the minimum inhibiting concentration in the medium and thus lowers their biocidal power.
65The study performed at cellular level by Obeidou in 1992 showed that the quaternary ammonium salts act on the enzymes of nitrogen’s primary metabolism. Furthermore, Nagal et al (1996) worked on seven strains of Pseudomonas fluorescens, one strain of Pseudomonas aeruginosa and one strain of Escherichia coli. The authors concluded that the strains tested are not capable of breaking down quaternary ammonium. No growth of bacteria was observed on a minimal medium with added quaternary ammonium as the only source of carbon and/or nitrogen. The results obtained specifically with the organisms from Lascaux Cave are therefore in agreement with those already published in the literature.
9.6 The contamination in 2007
66In July 2007, new black colonies appeared on the vaults and walls of the Passageway and of the Nave as well as on the decorated areas in the Apse. Samples were taken using swabs, and the analysis results show that these were moderately hydrophilic species (aw < 0.9) of Ulocladium sp. and Gliocladium sp. (fig. 133, 134).

FIG. 133 – Black colonies on the ceiling of the Apse.
© LRMH.

FIG. 134 – Ulocladium sp. (G x 400). • Ulocladium sp. (M x 400).
© LRMH.
9.6.1 First treatment
67Given the mechanical and chemical impact of the molds on the paintings and engravings, it once again seemed essential to proceed quickly with the biocidal treatment of all the contaminated zones and elimination of the mycelian residues. The aim of biocidal treatment is not to create a biological vacuum but, rather, to eliminate the visible organic mass, the sign of imbalance. This treatment must include the application of a chemical to eradicate the molds, followed by mechanical elimination to remove all the organic residues from the support. The purpose of the actual chemical treatment is therefore to reduce the reservoir of contaminants and stop the progression and extension of the fungi. It was also decided by the committee to put the cavity at rest for a period of at least three months.
68A new series of antibiograms was therefore scheduled. Six biocidal solutions, all based on quaternary ammonium, were tested to determine their effectiveness on isolated molds. Their composition was as follows:
– solution A: alkyl dimethyl benzyl ammonium chloride 50%;
– solution B: alkyl dimethyl benzyl ammonium chloride 25%, 2‑octyl‑2H‑isothiazole‑3‑one 2.8%, 2‑2‑oxydiethanol 25 to 50%;
– solution C: benzododecinium chloride < 2.5%;
– solution D: benzododecinium chloride 10% to 25%, miristalkonium chloride 10% to 25%, 2 octyl‑2H‑isothiazole‑3‑one < 2.5%;
– solution E: N formides, octylisothialozinone 5%.
69The results of the antibiograms show that solutions B, C and D perform the best. However the measurement of the solutions’ pH indicates an acidity that is too high in the case of solution B and this could represent a danger for the limestone support. The other two products chosen have a pH close to neutral and can be used without any risk. We chose solution D because it contains isothiazolinone, a fungicidal principle that strengthens the quaternary ammonium’s biocidal action. The treatment was applied by fogging (fine spraying in the form of mist) solution D locally on the contaminated surfaces for three consecutive days, at a rate of one application a day.
70A monitoring protocol was put in place to determine the treatment’s on‑site effectiveness. This protocol included various levels of verification:
1 – check of atmospheric contamination;
2 – observation of surface condition and photographic survey;
3 – measurements of overall metabolic activity by means of the adenosine triphosphate (ATP) which is the viability marker for organisms (Lundin et al. 1986); the ATP assay is based on the evaluation of the amount of light emitted; the intensity of the light emitted is proportional to the concentration of ATP; the emission of light, measured by means of a luminometer, is expressed in RLU (relative light unity).
4 – identification of the fungal species;
5 – evaluation of the viability of the microorganisms by means of epifluorescence microscopy; this technique is based on the use of a fluorescent dye which highlights the viable microorganisms (Jones, Senft 1985).
71All of these measurements were made before, during and after the treatment and they are expected to make it possible to monitor changes in the effects and effectiveness of the treatment. Eleven zones were chosen for the surface verifications.
72In the Passageway:
1 – black colonization on the left‑hand wall halfway along the Passageway (fig. 135);
2 – black colonization on the right‑hand wall opposite zone 1;
3 – white blanket on the ceiling halfway along the Passageway;
4 – black colonization on the left‑hand wall at the end of the Passageway;
5 – zone without down or visible colonization, left‑hand wall halfway along the Passageway;
In the Apse:
6 – black colonization on the left‑hand wall at the edge of the Shaft;
7 – black colonization on the right‑hand wall at the edge of the Shaft (fig. 136);
8 – black colonization at the bottom of the Apse beside the opening into the Shaft; In the Nave:
9 – black colonization beside the ‘Imprint’ panel;
10 – black colonization on the ceiling at the entrance to the Nave;
11 – zone without blanket or visible colonization on the wall, left‑hand side at the entrance to the Nave.

FIG. 135 – Appearance of the black colonies in zone 1.
© LRMH.

FIG. 136 –
Appearance of the colonies in zone 7.
© LRMH.
73After chemical treatment of the surfaces and after the cavity had been rested, the level of atmospheric pollution was, on the whole, low. At most of the measurement points, the mold rate did not exceed 67 CFU/m3 of air and the bacteria rate 84 CFU/m3 of air. No new species were identified in the air and some spores of Verticillium sp., Ulocladium sp. and Aspergillus sp. were found. At most points, the measurements of the overall metabolic activity before and after three months’ rest showed a decrease except at two points: test zone 6 and test zone 10.
74For test zone 10, an increase of the order of 17% was noted; in zone 6, however, an increase of the order of 50% was recorded. This increase can be explained either by the difficulty of accessing and treating all the zones uniformly (zone 10), or by the presence of cave‑dwelling animals: these organisms’ ATP interferes with that of the microorganisms (zone 6). Likewise, the large standard deviation between the results for the samples taken in this zone denotes the large disparity of the values, which was also linked to the presence or absence of insects on the sampled surface (fig. 137).

FIG. 137 – Comparison of pretreatment metabolic activity in December 2007 and after three months of rest, in March 2008.
© LRMH.
75The epifluorescence observations made before and three months after treatment also show a clear decrease in the number of viable fungal cells (little or no fluorescence) for most of the samples taken by adhesive tape (fig. 138a‑b). In test zone 6, certain fluorescent elements observed after treatment correspond to the debris from the cave‑dwelling animals still viable at the time the samples were taken (fig. 138c).

FIG. 138 – a zone 1, fluorescent fungal cells observed on the left wall in the middle of the Passageway in December 2007; b three months after treatment, no fluorescence was observed; c zone 6, debris of cave dwelling animals revealed by epifluorescence on the left wall at the level of the edge of the Shaft, three months after treatment.
© LRMH.
76Lastly, the comparison between the cultivated species indicates that there had been no change in the genera between December 2007 and March 2008. For certain zones, no fungi developed on the culture medium. Overall, there are still viable spores present despite a fall in the overall metabolic activity. The treatment monitoring results show that there is agreement between the ATP measurements, the epifluorescence observations and the swab‑based cultures. The atmospheric pollution had fallen notably after the cavity had been put to rest and, on the whole, the level of pollution was low.
77The treatment had a positive, but not radical effect on most of the zones. In the zones that were most contaminated (Passageway) and/or most difficult to access (vault of the Nave), this effect was, however, insufficient. The biocidal treatment had effectively played a role in the stabilization of the visible fungal contamination, but this stability remains fragile and additional treatment was proposed in order to reduce further the reservoirs of contaminants. The elimination of the overlaps of mycelia would seem to be essential in all the contaminated undecorated zones and in the decorated zones that can withstand it. Furthermore and because of the inaccessibility of certain zones (such as the vault of the Nave) spraying with a biocidal solution would be desirable, mechanical elimination being impossible at the current time.
9.6.2 Impact study
78In April 2008, a debate was initiated in the scientific committee concerning the appropriateness of pursuing the recommended treatments. It was therefore decided to set up another impact study of the potential treatments (chemical and cleaning) for the supports. A protocol was developed with criteria for selecting the test and cleaning and disinfection zones to be used to assess the treatment. Four zones (identified A, B, C, D) were selected in the Passageway according to the alterations of the walls and the level of microbiological contamination (fig. 139). There was specific treatment for each zone (fig. 140‑143). The cleaning and biocidal application operations were implemented by the team of restorers:
– cleaning with brushes with recovery of the residues;
– local application of a solution based on quaternary ammonium and isothiazolinone, completed on a small test area with an additional application of isothiazolinone alone.

FIG. 139 – Location of the four zones selected in the Passageway. Fond graphique :
© N. Aujoulat /ministère de la Culture et de la Communication, CNP.

FIG. 140 – Zone A, fragile zone with severe colonisation: gentle cleaning and biocide treatment.
© LRMH.

FIG. 141 – Zone B, fragile zone with black colonies: gentle cleaning without biocide treatment.
© LRMH.

FIG. 142 – Zone C, zone with severe black colonies: deep cleaning and biocide treatment.
© LRMH.

FIG. 143 – Zone D, zone with mild black colonies: gentle cleaning without biocide treatment.
© LRMH.
79The protocol for evaluating the levels of infection and the metabolic activity of the active microorganisms, before and after treatment, was obtained using several techniques, either by taking samples or by direct examination:
– surface swabs for the taxonomic study after growing on suitable media in the laboratory;
– surface swabs to determine metabolic activity by measurement of the ATP;
– microscopic examination of the contaminated surfaces using a portable microscope;
– analysis of the biogenic pigments by reemitted‑positron spectroscopy.
9.6.2.1 Culturing the swab samples
80The taxonomic study based on the swab samples taken from each zone shows the majority presence of Ulocladium sp. and Curvularia sp. sometimes accompanied by Verticillium sp. and Penicillium sp. The cultures show that no colonies developed from the samples taken from zones A and C which were treated with the biocidal solution (fig. 144).

FIG. 144 –Summary table of cultures before and after intervention
9.6.2.2 Culturing of the cleaning residues
81The residues recovered during the cleaning operation were analyzed to evidence their level of contamination. The microscopic examination makes it possible to note the significant overlapping of the mycelian filaments around mineral particles (fig. 145). 100 μl of a suspension prepared from those residues were inoculated on three synthetic media specific to the culture of the molds, DRBC (rose Bengal agar), CzD (Czapeck‑Dox) and DG18 (glycerol medium for xerophilous species). After 24 days’ incubation at 24 °C, the species were identified according to their morphological criteria and by means of molecular biology (fig. 146). On the whole, the isolates belong to moderately hydrophilic species such as Ulocladium sp., Curvularia sp., Cladosporium sp. and Ochroconis sp.

FIG. 145 – View of mycelian filaments surrounding the mineral particles.
© LRMH.

FIG. 146 – Cultures obtained from zone A, B, C, D residue.
© LRMH.
9.6.2.3 Additional test on the effectiveness of the biocides
82By diffusion in synthetic culture media
83Three different culture media (DBRC, CzD and malt agar) were inoculated with the solution based on the residues used previously. This action was repeated five times for the repeatability of the results. One day later, a hole with a diameter of 1 cm was made in the center of the seeded culture medium and this space was filled with various biocides:
– 5% mixture of quaternary ammonium and isothiazolinone (Dévor Mousse),
– 3% solution of isothiazolinone (Parmétol DF12) [fig. 147],
– ready‑to‑use solution of a mixture of isothiazolinone and quaternary ammonium (Keim Algizid Plus),
– Dévor Mousse + Parmétol DF12 mixture,
– Dévor Mousse + Keim Algzid Plus mixture.

FIG. 147 – Evidence of the fungicidal action of Parmétol DF12 after 16 weeks.
© LRMH.
84Macroscopic observation of the different media shows that Parmétol DF12 improves the effectiveness of the biocidal treatment. However, the mixture with the quaternary ammoniums can be recommended to facilitate the penetration of the fungicide – such as isothiazolinone – into the fungal biofilm and decrease any possible allergic impact of treatment with isothiazolinone alone.
85On clays
86The effectiveness of the biocidal formulations was also tested directly on clay sediments taken from Lascaux. After having mixed these clays with water, the paste obtained was placed in plastic boxes. After solidification, these supports were inoculated with the strains taken directly from the cave. The specimens thus prepared were incubated on a bed of vermiculite in enclosures simulating the high humidity present in the cavity. After two weeks, the surfaces of the contaminated clays were treated with the various biocidal products including hydrogen peroxide (H2O2). Whereas the treatment with H2O2 visibly made the clay material swell immediately, the application of the quaternary ammonium fungicide plus isothiazolinone formulation gave the best results for verification of re‑contamination of the clay soil.
9.6.2.4 Dévor Mousse mineralization test
87This test is performed using Biolog MT microtiterplates. It consists of seeding micro‑recesses (in this case 96) prepared with a redox indicator that changes color according to microbial activity. One series of recesses is filled with a solution of glucose at concentrations decreasing by a factor of 10, starting from 10% (10%, 1%, 0.1%, etc.). The other series was filled under the same conditions, with a solution of Dévor Mousse. The recesses were then seeded with a mixture of microorganisms taken from the cave’s clay sediment samples. The results clearly show that microbial activity is stimulated with high glucose concentrations. With the Dévor Mousse, the microbial activity was inhibited starting from a concentration lower than 0.1% and increased very slightly at 0.01% (fig. 148).

FIG. 148 – Biolog MT plates with decreasing glucose concentrations (left) and Dévor Mousse concentrations (right).
© LRMH.
88The results of these tests with the MT plates, widely known as a tool for revealing the biodegradation of organic compounds, make it possible to conclude that the Dévor Mousse is not degraded by the microorganisms in the cavity.
9.6.2.5 Measurement of metabolic activity
89The measurements taken before treatment, after the first cleaning test, after actual cleaning and then, lastly, after the biocidal treatments indicate a notable fall in metabolic activity after treatment with the biocidal solution and an absence of activity in the zones that received further treatment with isothiazolinone alone (Parmétol DF12) [fig. 149].

FIG. 149 – Comparison of ATP measurement results from the four zones.
© LRMH.
9.6.2.6 Results
90The latest results of the microbiological impact study show that treatment with a mixture of quaternary ammonium and isothiazolinone followed by an application of isothiazolinone alone gives very good results. Five months after treatment, there was no visible recurrence of colonization in these zones. However, the cleaning impact study shows that in the right‑hand part it is very difficult to work on the supports, which are very fragile and have a large number of engravings.
9.7 Conclusions
91Today, even if the contamination is no longer comparable, in terms of intensity, with what we saw at the beginning of the crisis in September 2001, it has still not been eradicated. Everything leads us to believe that the cave’s equilibrium is precarious, and to restore it we must know the exact environmental parameters that initiated the developments in order to contain them. In all treatment operations, the end result is influenced by four interdependent factors, grouped in the factorial circle, called the “Sinner circle”. If we adapt this circle to the problem of managing a biological crisis, these factors are as follows (fig. 150): – climate management, – cleaning, – biocidal treatments, – repeated action. However, in a comprehensive intervention strategy, if one of the factors is reduced, it must be compensated for by increasing one or more of the other factors in order to to maintain the final quality of the result. In the case of the right‑hand part of the cave, there must be discussion of the right balance between the different actions. The gradual advances in our knowledge of the chemical, climatic and microbiological environment should eventually make it possible to define the best strategy for conserving the works, while respecting this particularly complex milieu.

FIG. 150 – Sinner Circle.
© LRMH.
92References
93Brunet, Vouvé 1996 : BRUNET (J.), VOUVÉ (J.). — La conservation des grottes ornées. Paris : CNRS, 1996, 263 p.
94Dupont et al. 2007 : DUPONT (J.), JACQUET (C.), DENNETIERE (B.), LACOSTE (S.), BOUSTA (F.), ORIAL (G.), CRUAUD (C.), COULOUX (A.), ROQUEBERT (M.‑F.). — Invasion of the French Paleolithic painted cave of Lascaux by members of the Fusarium solani species complex. Mycologia, 99, 4, 2007, p. 526‑533.
95Isomaa 1984 : ISOMAA (B.). — Membrane‑perturbing and lytic properties of surface‑active alkyltrimethylammonium salts. Ecological bulletins, 36, 1984, p. 26‑30.
96Jones, Senft 1985 : JONES (K.), SENFT (J.A.). — An Improved Method to Determine Cell Viability by Simultaneous Staining with Fluorescein Diacetate‑Propidium Iodide. The journal of histochemistry and cytochemistry, 33, 1, 1985, p. 77‑79.
97Lefèvre 1974 : LEFÈVRE (M.). — La maladie verte de Lascaux. Studies in conservation, 19, 3, 1974, p. 126‑156.
98Lefèvre, Laporte 1969 : LEFÈVRE (M.), LAPORTE (G.‑S.). — The ‘maladie verte’ of Lascaux : diagnosis and treatment. Studies in speleology, 2, 1, July 1969, p. 35‑44.
99Lundin et al. 1986 : LUNDIN (A.), HASENSON (M.), PERSSON (J.), POUSETTE (A.). — Estimation of biomass in growing cell lines by Adenosine Triphosphate assay. Methods in enzymology, 1986, 133, p. 27‑42.
100Nagal et al. 1996 : NAGAL (K.), OHTA (S.), ZENDA (H.), MATSUMOTO (H.), MAKINO (M.). — Biochemical Characterization of Pseudomonas fluorescens Strain Isolated from a Benzalkonium Chloride Solution. Biological and pharmaceutical bulletin, 19, 6, 1996, p. 873‑875.
101Obeidou 1992 : OBEIDOU (W.). — Influence de quelques sels d’ammonium quaternaire sur la croissance et le métabolisme azoté de divers organismes : champignons du bois et noyer. Thèse doct., univ. Nancy 1, 1992. 114 p.
102Orial, Mertz 2006a : ORIAL (G.), MERTZ (J.‑D.). — Lascaux : une grotte vivante : étude et suivi des phénomènes microbiologiques. Monumental, 2, 2006, p. 76‑87.
103Petersen et al. 1993 : PETERSEN (K.), KRUBEIN (W.E.), HÄFNER (N.), LUX (E.), MIETH (A.). — Aspects of biocide application on wall‑paintings : report on eurocar project EU 489, Biodecay. In THIEL (M.J.) ed. — Conservation of stone and other materials, vol. 2. London : Spon, 1993, p. 597‑604.
104Roquebert, Orial 2002 : ROQUEBERT (M.‑F.), ORIAL (G.). — Analyse d’une contamination : comment analyser une contamination biologique et en évaluer les risques. In ROQUEBERT (M.‑F.). — Les contaminants biologiques des biens culturels. Amsterdam : Elsevier ; Paris : Muséum national d’histoire naturelle, 2002, p. 147‑159.
105Salton 1951 : SALTON (M.R.J.). — The adsorption of Cetyltrimethylammonium bromide by bacteria, its action in releasing cellular constituant and its bactericidal effect. Journal of general microbiology, 5, 1951, p. 391‑404.
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