Non-human biological agents involved in the deterioration of rock art panels range from large mammals through to microfauna, from large plants through to cryptogamic growths, and include bacteria. All rock surfaces bear biotic substances, and all rock surfaces contain organic carbon (Bednarik 1979), which is the principal reason why carbon dating of rock surface features only yields questionable results in many circumstances. Rock surfaces are sometimes implicitly regarded as somehow ‘sterile’, when in fact bacteria occur even several kilometres deep in the lithosphere. Micro-organisms are also implicated in the formation and recycling of specific accretionary deposits as well as in the removal of various cations from numerous mineral components. Not only do such activities contribute to the weakening of rock fabric (Berthelin 1988), they are contingent upon the production of corrosive agents. Their metabolic products may also be corrosive (Eckhardt 1978). Broadly speaking, we are concerned with bacteria utilising atmospheric nitrogen (converting it to ammonia, or oxidising this to nitrous or nitric acid); those oxidising sulphur to sulphur dioxide and converting it to highly corrosive sulphuric acid or the weaker sulphurous acid; those oxidising iron and manganese from minerals; and those attacking silicates and phosphates with organic acids, in particular ketone-based compounds. However, two points need to be considered here: first, the quantities of such highly effective acids are extremely small; second, the processes they cause are not necessarily only corrosive. For instance Ferrobacillus sp. or Thiobacillus sp. do consume cations, but often from airborne material rather than the host rock, and they may accrete protective surface crusts that actually preserve petroglyphs. Similarly, the mobilisation of amorphous silica, for instance through 2-ketogluconic acid produced by bacteria, can lead to the deposition of silica skins, which have been observed to assist in consolidating both petroglyphs and pictograms. Therefore microbial action is not readily identifiable in most cases, nor are the processes readily definable. An exception is the effect of bacteria on some organic rock art paint residues, such as those reported by me from India (Bednarik 1992a: 148). At some sites in Madhya Pradesh, a white pigment is systematically transformed into black, generally from the margins of the coloured areas inwards. Similar bacterial action may occasionally account for other significant colour changes in paint residues (e.g. McNickle 1991).

The second group of Thallophyta, the fungi, also require no chlorophyll, they too can exist in complete darkness. Small airborne organic matter and moisture is all that is needed by many species. By producing oxalic or citric acid, some species can attack various minerals more aggressively than bacteria, except calcium-rich minerals which seem to be more susceptible to bacterial attack. The citric acid of fungi lead to hydrolysis of silicate minerals (Silverman and Munoz 1970; Winkler 1973). The only fungi (moulds) of interest to rock art science are saprotrophs, i.e. those feeding on dead organic matter, and they occur commonly in limestone caves and at other locations of high humidity. Fungi are further of interest as the symbiotic partners of algae in the formation of lichens, in which the latter provide atmospheric carbon and nitrogen through photosynthesis, while the fungi contribute water and minerals.

In contrast to fungi, algae require light for photosynthesis, although filtered and weak light suffices for some species. While algae range from single-celled forms to very complex multicellular seaweeds, two types are of particular interest to the rock art conservator. The blue-green algae (Cyanophyta), because of their involvement in the production of rock varnish (Scheffer et al. 1963), and the terrestrial green algae (Chlorophyta), which like the former class occur in lichen. Algae do not seem to attack rock surfaces directly, but they can have a damaging effect through their ability of moisture retention, and they depend on the presence of moisture. Especially on carbonates, this may effect minor solution in the presence of carbon dioxide. Algal growth can be unsightly when it occurs on or near rock art but is probably not a significant conservation threat.

Lichens, as already noted, result from a symbiosis between fungi, usually of the class Ascomycetes, and blue-green or green algae. They have been a favourite target in rock art conservation work, but there is very limited evidence of real damage lichens have caused to petroglyphs. Many reports appeal to issues of ‘aesthetics’, and most empirical claims appear to be anecdotic. Some authors even suggest that a lichen cover may reduce rock weathering rates (Wilson 1995), or while acknowledging deleterious effects have warned that removal of lichen may exacerbate substrate deterioration (Florian 1978). Jackson and Keller’s (1970) study, while showing that the weathering zone under lichen cover is thicker than on nearby rock that is free of lichen, is inconclusive, because it cannot tell us whether this is because the lichen has prevented erosion, or because the weathering zone develops faster under the lichen (which I regard as unlikely). Cochrane and Berner (1992, 1993) report the preservation of a silica skin beneath colonising lichens, while other authors suggest that lichens could enhance rates of weathering (Fry 1924, 1927; Syers and Iskandar 1973). The significantly conflicting opinions of various investigators are perhaps easily explained: there are many lichen species to be considered, and it would be surprising if they all had similar effects on component minerals. By the same token each species certainly affects each rock type differently. So until we have acquired a great deal more knowledge about the effects of various lichens on various rock types we need to acknowledge that the jury is still out on lichens, and that it would be premature to favour lichen removal. On balance, the evidence currently available suggests that the principal problem with lichens is that they retain moisture which may hasten weathering, but they may also stabilise surfaces on some types of rock, and their chemical weathering impact on most rock types is probably negligible.

Mosses have attracted considerably less ire from rock art site managers, although they are no doubt more effective in retaining moisture for extended periods of time and thus probably a greater conservation threat. Mosses are non-flowering plants that lack roots, but whose stems are equipped with rhizoids to anchor them. A rhizoid is a hairlike filament used to secure the plant to its substrate and it absorbs water and nutrients, operating like roots but of considerably simpler construction. Rhizoids are also found in fungi and lichens. Because of their often much greater bulk in comparison to lichens, mosses attract airborne debris and edaphic sediment to assist in colonising rock surfaces, which understandably conflicts with the survival of rock art.

Even larger plants, such as Gramineae, shrubs and trees can become major agents of rock weathering. In Chapter 3, two basic types of rock markings by plants were distinguished, ‘kinetic’ and ‘chemical’ marks. The first include those occasioned by large tufts of grass or plant branches rubbing against relatively soft rock surfaces over years; and marks made by tree roots hugging rocks for support, and moving slightly as the tree sways in the wind. The second type of marks, the ‘chemical’ marks, is particularly prominent on carbonate rocks, and is attributable to mycorrhizal action along the roots of plants, i.e. to respiratory carbon dioxide and possibly organic acids secreted by the mycelia of symbiotic fungi. While in some exceptional cases such rock markings resemble petroglyphs so closely that archaeologists have misidentified them as such, the influence of plants in the breakdown of rocks is much broader. It includes the widening of fractures with roots, the deposition of organic litter, the shade provided to encourage the colonisation of rocks by cryptogamic organisms, and the destruction or rock surfaces by brushfires.

There are well-known threats to rock art by several insect species, particularly mud-daubing wasps and termites. In Australia alone, four species of termites (Isoptera) have been reported to be causing damage to rock paintings through their construction of ‘runways’ over vertical walls (Watson and Flood 1987). These tunnel-shaped tracks connect the central nest with feeding areas, protecting the insects from predators and desiccation. Termites are an order of social insects that occur widely in tropical and subtropical regions and feed on complex carbohydrate cellulose. I have observed their runways, as they are preferably called, on rock art panels in numerous countries. These runways consist of an arched sheathing of soil, rock fragments and digested wood, and they are only built by a minority of the many termite species. Occasionally even the nest or mound structures are located at art sites, sometimes concealing part of a rock wall. The main damage, however, is through the runways radiating from the nest. In their construction, sand grains are removed from a supporting sandstone panel and placed on the surface of the runway sheath. In addition to this form of biological weathering, when the runways eventually exfoliate, any paint residues they cross are likely to flake off as well, leading to rapid destruction of the affected rock art.

Mud-daubing wasps (Hymenoptera) occur commonly in most if not all continents, and the nests of numerous species have been observed at rock art sites. A great variety of different structures (e.g. tubular, sub-hemispherical, polygonal) can even leave pigmented imprints on art panels after their exfoliation. Where bonding of the nest structure by salt precipitation has occurred, exfoliation is likely to remove some of the host rock’s surface. The nests consist of locally derived fine sediment and organic ketones. There is no evidence that these insects erode the rock surface, therefore the damage they inflict on rock art is limited to the exfoliation phase of their nests, but as they tend to attract more nest-building activity the result can be unsightly colonies of such ultimately damaging structures (Naumann and Watson 1987). Bees can damage rock art in one of two ways. Many sites contain bee hives made essentially of wax, but some species also burrow into soft rock types, such as calcarenite or weathered sandstone. Such nesting aggregations can include hundreds of closely adjacent entrances, each several millimetres in diameter.

The damage of mud nests to rock art is not restricted to those of insects. Sallows and martins often build their nests of mud pellets in rockshelters and limestone caves (Bednarik 1988a). Birds also damage rock art through the acidity of their droppings, which facilitates dissolving of rock varnish and other protective deposits (Bednarik 1979) and estab-lishment of cryptogamic growths. In regions of deeply patinated rock outcrops, the highest points are inevitably free of patination, because they are the favourite perches of birds of prey. Therefore discolouration of such an aspect is not necessarily evidence of lightning strike.

Numerous larger animals visit rock art sites and damage the art in the process. In Australia, for instance, feral pigs, water buffaloes and domestic as well as feral cattle are among the main culprits. Hundreds of sites have been fenced in to keep these animals from rubbing their bodies against the walls of the rockshelters, and from raising dust which settles on the art. Perhaps the greatest problem with these animals is that many suffer from a variety of parasites which prompt them to rub parts of their bodies vigor-ously against the walls of rockshelters, whose relative coolness they are attracted to in a hot climate. This has devastating effects on any rock art present. Similar or other mammalian species affect rock art, especially pictograms, in most parts of the world, and while indigenous species do contribute to this damage, the impact of introduced species, both domesticated and feral varieties, is usually far greater. This is primarily because native megafauna has now been largely eliminated around the world, but on rock surfaces that have preserved earlier traces, particularly in deep caves, the picture is different. In the caves of Europe, the most prominent rock markings are not those of humans, but those of megafauna, particularly of Europe’s cave bear. They consist not only of extensive scratch marks but also of the Bärenschliffe, polished walls where countless gen-rations of bears polished cave walls with their bodies (see ‘Animal markings’). Animal markings occur in most caves and are a form of biological weathering, as are rock surface modifications by humans, such as bulldozer marks, mining quarries or rock art. Anthropogenic weathering, however, is so extensive that it warrants a separate section (see below).

REFERENCES Bibliography of Rock Art Conservation