In vitro investigations of the impact of different temperature and flow velocity conditions on tufa microfabric - Tufas and Speleothems: Unravelling the Microbial and Physical Controls - page 201

by the time that these aggregates reached 2-10 mm
in size many had developed well-formed rhombic
crystal faces by tabular stacking of the nanospheru-
lites (see Fig. 10a, b; Pedley et al. 2009). Continued
crystal growth, and further reduction of accommo-
dation space within the EPS, ultimately led to the
development of a competitive growth fabric in
most basal calcite layers. Interestingly, the striated
compromise boundary surfaces between adjacent
crystals also show that tabular nanospherulite stack-
ing lies parallel to these striations (Fig. 10c, d).
Ultimately, many crystals bonded with the flume
substrate beneath the EPS. However, some biofilm
areas of the basal carbonate layer failed to attach
and the floating basal calcite layer developed
rhombic crystal terminations at the base of the crys-
tals (cf. Fig. 11a with Fig. 4b). Similarly, the floating
calcite microspar crystals within the EPS septal
zones also developed free crystal faces (Fig. 11b)
although a competitive growth fabric did develop
towards the base of the septae and eventually
buried the upper surface of the basal layer (see
Fig. 4a). Despite the floating nature of crystal
growth it is clear that the majority of crystals grew
with sub-parallel c-axes which were all orientated
normal to the flume surface even though many
never became physically attached to it.
This process raises a number of fundamental
issues related to the development of crystals
within freshwater EPS. From the previous discus-
sion it is clear that initially, the cyanobacterial and
algal filaments were the only elements within the
biofilm to be anchored to and orientated normal to
substrate. Subsequently,

zones, most easily defined by nanospherulite micro-
peloid development within the EPS, also developed
normal to biofilm base. Finally, the microspar crys-
tals which developed from these micropeloids also
developed long-axis alignment normal to the base
of the biofilm.
Clearly, the underlying solid substrate exerted
no direct control on the primary orientation of the
nanospherulite stacking patterns. Consequently,
the development of the subsequent calcite crystal
orientations cannot be entirely accounted for by
the 'heterogeneous nucleation' mechanism outlined
in Turner & Jones (2005). It is also unlikely that
cyanobacterial filaments exert any direct orientation
control (cf. template hypothesis of Monty & Mas
1979) because of the general lack of contact
between crystal and microbial sheath material, a
feature also observed in marine stromatolite mats
by Dupraz et al. (2004). However, coccoid (hetero-
trophic) bacteria are frequently present in surface
depressions on and within the precipitates (e.g.
Fig. 3d). In addition, there appeared to be domi-
nance of them in the vicinity of the polygonal
septae within the EPS. Equally significant was the
ramification of EPS throughout these embryonic
crystals.
Precipitation mechanisms driving crystal growth
are becoming better understood (see Rogerson et al.
2010). Briefly, studies of the biological influence
on diurnal variations in pH (Rogerson et al. 2008;
Liu et al. 2008; Shiraishi et al. 2008; Bissett et al.
2008) indicate that an excess of CO 2(aq) is generated
throughout the night as a metabolic byproduct of
respiration, and consequently promoted a low the polygonal septal

Fig. 11. (a) Lower surface of a basal calcite layer which never attached to the flume substrate during development.
Consequently, individual crystals within it have developed free rhombic faces by continued growth along their c-axes.
The localized absence of basal calcite layer attachment encouraged the development of an undulose upper surface to the
biofilm. Note that calcite rhombs at the base of the basal calcite layer have never been in direct contact with the
underlying flume surface. Autumn, fast-flow experiment. Air dried SEM sample. (b) An area of basal calcite layer
(attached surface to the left) examined immediately after being detached from the flume substrate. Note how freedom of
crystal growth on the upper surface of the basal calcite layer is now compromised by downward growth of larger spar
crystals within the overlying EPS cover (also with their c-axes orientated normal to substrate). When all EPS between
these two crystal sets has become occluded the crystal compromise boundaries will be indistinguishable from a
physico-chemically precipitated two generation cement mosaic grown directly on to a solid substrate.