Geoscience Reference
In-Depth Information
abundance of Mn, the presence of epilithic organisms, the
morphology of the underlying rock, aeolian abrasion and
the proximity of the varnish to soil (Dorn, 1986, 2007).
Rock varnish comprises a mixture of about two-
thirds clay minerals (commonly mixed-layer illite-
montmorillonite) cemented to the host rock by typically
one-quarter manganese and iron oxyhydroxides (birnes-
site and haematite, respectively), the remaining con-
stituents comprising minor and trace elements (Potter and
Rossman, 1979; Dorn and Oberlander, 1982). Varnish Mn
: Fe ratios vary from less than 1 : 1 to over 50 : 1, in con-
trast with Mn : Fe ratios of about 1 : 60 in the Earth's crust
(Dorn, 2007). In general, blacker varnishes contain more
manganese (typically 20 % MnO
2
by weight) while iron
predominates in orange varnish (10 % FeO
2
by weight
and less than 3 % MnO
2
; see Potter and Rossman, 1977;
Perry and Adams, 1978; Elvidge and Moore, 1979). In
addition to Si, Al, Mn and Fe, major elements found in
most varnishes include K, Na and Ti. The proportion of
other elements varies considerably. Ba and Sr are the most
abundant trace elements, but Cu, Ni, Zr, Pb, V, Co, La,
Y, B, Cr, Sc and Yb, in order of decreasing concentra-
tion, are found in all varnishes (Engel and Sharp, 1958;
Lakin
et al.
, 1963). Trace heavy metals generally covary
with Mn abundance, and sometimes with Fe, due to the
scavenging properties of the Mn and Fe oxyhydroxides
(Thiagarajan and Lee, 2004; Tebo
et al.
, 2004).
Rock varnish may be of geomorphic importance in pro-
tecting rocks from weathering (Merrill, 1898). However,
Engel and Sharp (1958) found that varnish is readily de-
stroyed by mechanical flaking and dissolution under wet
conditions. Resistance to removal is related to chemistry,
clay mineral content and the roughness of the underly-
ing substrate, with Si-rich varnish on originally rough
substrate having the highest resistance and some Fe-rich
varnishes on smooth rock being sufficiently soft to rub
off with a fingertip (Oberlander, 1994). Dorn and Ober-
lander (1982) reported hardnesses approaching 6.5 Moh
and Allen (1978) held that in the Sonoran Desert it is
destroyed only by sandblasting.
Any coherent explanation for the origin of rock var-
nish must be tied to, and explained by, the processes
that lead to (a) the enhancement of Mn concentrations
on surfaces and (b) the fixation of clay minerals by Mn
and Fe oxyhydroxides to such surfaces (von Humboldt,
1812; Lucas, 1905; Engels and Sharp, 1958). Theories
of varnish development can be divided into two groups,
according to whether the coating formed from an internal
or external source (Drake, Heydeman and White, 1993).
Much early work suggested that varnishes developed as
moisture was drawn out of rocks, precipitating the min-
1901; Blake, 1905). However, this process of 'sweating'
would be expected to form a leached and weathered zone
beneath the varnish. Occasionally this is present (Allen,
1978; Glasby
et al.
, 1981) but the coating is technically
then a weathering rind.
The consensus today is that the majority of the con-
stituents that make up a rock varnish are derived from
allochthonous sources (Allen, 1978; Perry and Adams,
1978; Elvidge and Moore, 1979). Such sources help ex-
plain the presence of trace elements that are absent from
the host rock (Lakin
et al.
, 1963; Knauss and Ku, 1980).
Four conceptual models have been proposed to explain the
build-up of a varnish over time. The first invokes abiotic
surficial chemical weathering to increase Mn : Fe ratios
(Linck, 1901; Engel and Sharp, 1958; Marshall, 1962;
Whalley, 1984; Smith and Whalley, 1988, 1989). Under
this model, small pH/Eh fluctuations towards more acid
conditions dissolve Mn but not Fe (Krauskopf, 1957). The
released Mn is then fixed in clays following the evapora-
tion of surface water or a further change in pH. Dorn
(1989, 2007, 2009) presents a detailed critique of this
model, identifying a number of features of varnishes that
are inconsistent with a purely abiotic origin. First, in ad-
dition to occurring in deserts, varnishes are found in en-
vironments that are too wet and acidic to oxidise Mn
(Dorn, 1998). Second, varnishes found in wetter climates
are typically higher in Mn than those in arid, more al-
kaline, settings (Dorn, 1990), a result that would not be
predicted by the high pH requirements of abiotic oxi-
dation. Third, Mn-rich rock varnish is not common in
some environments, such as coastal fog deserts, where re-
peated pH/Eh fluctuations might be expected to enhance
Mn levels. Fourth, given that dust deposition and pH/Eh
fluctuations occur at least annually in all but the most arid
environments, varnishes should accrete many orders of
magnitude faster than is seen in nature (Liu and Broecker,
2000, 2007). Overall, while abiotic processes may be in-
volved in varnish formation (Bao, Thiemens and Heine,
2001), including clay cementation (Potter and Rossman,
1977; Dorn, 1998) and trace element enhancement dur-
ing wetting (Thiagarajan and Lee, 2004), these processes
alone would not generate a rock varnish.
A second conceptual model proposes that microorganic
processes generate and bind the constituents that produce
rock varnish; studies in North Africa (Drake, Heydeman
and White, 1993), North America (e.g. Dorn and
Oberlander, 1981, 1982; Palmer
et al.
, 1985; Nagy
et al.
,
1991), South America (Jones, 1991), Australia (Staley
et al.
, 1983; Dorn and Dragovich, 1990) and the east-
ern Mediterranean (Krumbein and Jens, 1981; Hungate
et al.
, 1987) all support such an origin. A wide range
