Geoscience Reference
In-Depth Information
darker anatase or iron oxide staining, and have either a
concentrically zoned or undifferentiated internal fabric.
Glaebules are often used as diagnostic indicators of pedo-
genesis. However, some glaebular structures within Kala-
hari silcretes formed at or near the watertable (Shaw and
Nash, 1998) or were inherited by silica replacement of
pedogenic calcretes (Nash, Thomas and Shaw, 1994), so
caution is needed. It is also important to ascertain whether
glaebules formed in situ during silicification or had a de-
trital origin (Hill, Eggleton and Taylor, 2003).
Void spaces within silcretes may be filled by a variety
of minerals (Figure 8.13(c) and (d)). The most common
sequence is for massive or laminated opal to line void mar-
gins, followed by layers of chalcedony, microquartz and
megaquartz (Summerfield, 1982; Thiry and Milnes, 1991;
Rodas
et al.
, 1994). The increasing organisation of silica
minerals towards the centres of voids may be accompanied
by fractionation of silica isotopes, with Basile-Doelsch,
Meunier and Parron (2005) reporting
undissociated monosilicic acid (either as the monomer
H
4
SiO
4
or the dimer H
6
Si
2
O
7
; see Dove and Rimstidt,
1994). Waterborne transfer mechanisms can be divided
into lateral or vertical movements, although both can oc-
cur in combination. Lateral transfers are of importance
for the development of nonpedogenic silcretes (Stephens,
1971; Hutton
et al.
, 1972; Hutton, Twidale, Milnes, 1978).
Silica transfer in river water, for example, is a key control
upon the formation of drainage-line silcretes (Shaw and
Nash, 1998), while pan/lacustrine silcrete development
is dependent upon the movement of dissolved silica into
topographic lows. Subsurface lateral water transfers are
vital for groundwater silcrete formation and may provide
a silica source for other silcrete types. Vertical transfer
mechanisms involve movements of silica-rich solutions
supplied from surface waters, leaching of overlying sedi-
ments (S. H. Watts, 1978a, 1978b; Marker, McFarlane and
Wormald, 2002) or aeolian or biological sources (Goudie,
1973). Dissolved silica may be washed down through a
soil or sediment or drawn upwards from the water table.
In pedogenic silcretes, downward-moving silica may be
supplied by dissolution during profile development, with
silica mobilisation and precipitation promoted by wetting
and drying respectively (Thiry, 1999).
Silica precipitation in near-surface environments is con-
trolled by the concentration of silica in solution and the
duration of wetting/drying cycles (Knauth, 1994). If a
solution is supersaturated with respect to amorphous sil-
ica, silica monomers may polymerise and aggregate to
form colloids, which may in turn precipitate as opal-A
(Williams, Parks and Crerar, 1985). Quartz may nucle-
ate from dilute solutions providing there are no inhibiting
minerals or ions present (chlorite, illite, haematite, Fe or
Mg), the solutions are slow moving and there is a template
such as a quartz grain to initiate deposition (e.g. Heald and
Larese, 1974; Ollier, 1978; Summerfield, 1982). Crystal
size and order are controlled by the speed of precipita-
tion, itself influenced by the host material permeability
and evaporation rate.
Silica precipitation is also influenced by pH, Eh, evap-
oration, the presence of other elements and minerals in
solution, biological life processes and, to a minor extent
in near-surface settings, temperature and pressure. Silica is
only weakly soluble in neutral pH water (10 ppm at 25
◦
C
for quartz; see Siever, 1972) with solubility increasing
significantly only at above pH 9.0 (Dove and Rimstidt,
1994). The solubility of amorphous silica reaches 800
ppm at pH 10.6 (Alexander, Heston and Iler, 1954); such
alkaline conditions are not unusual in saline lakes (see
Chapter 15). Following the dissolution of silicates at pH
9.5 to 10.5, desiccation can lower the pH to about 7.0,
δ
30
Si values as low
as
for megaquartz cements within groundwater
silcretes. The nature of the host material may also influ-
ence silica mineralogy (e.g. S. H. Watts, 1978a, 1978b;
Summerfield, 1982; Wopfner, 1983; Webb and Golding,
1998), with opal prevalent in clay-rich substrates and mi-
croquartz common in more porous sediments or carbon-
ates (Callen, 1983; Thiry and Ben Brahim, 1990; Benbow
et al.
, 1995). Void fills of other minerals such as calcite
(Summerfield, 1982, 1983c; Nash and Shaw, 1998), iron
oxides or zeolites (Terry and Evans, 1994) are also re-
ported.
−
5.7
8.6.4
Mode of formation
There is considerable debate concerning the mode of for-
mation of silcrete. This is not helped by the fact that, with
the exception of biogenic silcretes in Botswana (Shaw,
Cooke and Perry, 1990), duripans in North America (Flach
et al.
, 1969; Chadwick, Hendricks and Nettleton, 1989;
Dubroeucq and Thiry, 1994) and dorbanks in South Africa
(Ellis and Schloms, 1982), most are relict features. There
are a number of potential silica sources. Chemical weath-
ering of silicate minerals is the most important, although
replacement of quartz by carbonates may be significant in
some environments. Surface solution of quartz dust may
provide a source in deserts (Summerfield, 1983a), together
with biological inputs from silica-rich plants and microor-
ganisms. Silica may be transported to the site of silcrete ac-
cretion by two mechanisms. Quartz dust, plant phytoliths,
sponge spicules and diatoms may be blown considerable
distances by the wind. All other transfers rely upon the
