The lateritic bauxite in the Darling Range of Western Australia is approximately 4-6 m deep and is composed of caprock (duricrust) and underlying friable bauxite developed from highly weathered granite and dolerite. This paper investigates the impact of mine floor ripping operations on materials exposed after mining in relation to strength, water retention, and plant root growth. Deep ripping has been shown to create a structured rootzone and increase both recharge and plant available water. Material classified as quartz rich (Zm) lacks structural features for plant root growth, yet the material requires the least force to rip. These materials require deep preripping followed by contour or multi-tine ripping for successful restoration. Materials that contain 50% or more coarse fragments (gravels, cobbles, and stones) increase the force required for pre-ripping operations but only require contour or multi-tine ripping; doleritic clay with a high coarse fragment content may be readily colonized by roots without pre-ripping. Localized compression (smearing) of materials may occur during ripping due to the pre-ripping tine working deeper than its critical depth of 1.0-1.5 m and/or ripping of materials at high moisture content. A relationship exists between the ratio of reactive silica to total silica and the clay content of regolith, both properties being determined from pre-mining borehole samples. This ratio can be used to identify future mine floor materials from borehole data prior to mining and to target ripping operations to appropriate regolith types to enable the most effective and economical restoration practice.

The Mount Charlotte quartz vein gold deposit comprises a series of steeply plunging, pipelike vein stockwork orebodies in massive metagabbro. The orebodies are strata bound to the most differentiated unit of the host sill and are typically adjacent to major steeply dipping faults that cut the sill. The stockworks have two sets of veins with a dihedral angle of about 50°that developed as hydraulic fractures, filled simultaneously, and are generally approximately equally developed. Veins crosscut major faults and parallel minor faults of two sets but are cut along reactivated fault surfaces. The fault sets were inactive during mineralization and are neither veined nor are loci of zones of intense alteration. Rare faults of a third set are in part veined and are loci of zones of mineralization. Two interpretations of the stress regime during vein formation are based on different models of fracture formation. For both stress regimes, the major fault sets are relatively unfavorably oriented for slip or dilation, and predicted movement vectors do not fit fault-plane lineations. The lack of fault activity during ore fluid flow promoted formation of vein stockworks at Mount Charlotte rather than shear zone or fault-hosted veins. Fluid flow paths and orebody siting are controlled by stress-guide effects due to the rheology of the host gabbro, and by the three-dimensional geometry of impermeable faults and of fault-bounded blocks of rock.

No abstract

BACKGROUNDSome rehabilitated bauxite mines in southwestern Australia have accelerated gully erosion while the majority have little erosion. Anecdote suggests that gully erosion in rehabilitated forest is controlled by slope gradient but not all steep areas erode and conversely, some gentle slopes do. Mining companies aim to achieve erosion behaviour in rehabilitated areas similar to that of the surrounding forest -where large gullies are rare. We surveyed 26 eroding and erosion-prone rehabilitated hillslopes and developed descriptive models to predict the occurrence of gully erosion. A model of gully triggers implies that triggers and threshold effects are as influential as slope gradient and length in determining both the occurrence and severity of gully erosion. Many pre-existing triggers that predispose critical parts of a landscape to gully erosion activate only under threshold-excess conditions. Pinched concavities (thalwegs), shallow topsoil and gravel cover, erodible subsoil, high groundwater level, misplaced fauna habitats and irregular rehabilitation boundary edges are common erosion triggers. Slope angle and slope length, upslope catchment area, landscape position, soil storage and infiltration capacity, and rainfall (duration and intensity) are threshold variables. Many of these triggers and some thresholds can be identified and hence mitigated at the pre-mining stage. Topographic thresholds for gully erosion determined by the relationship between the critical slope (S cr ) and contributing area (A) at Boddington, Huntly and Willowdale bauxite mines are: S cr = 0.2A -0.39 , S cr = 0.05A -1.66 and S cr = 0.02A -1.59 . Additionally, at the minimum catchment area for gully incision (0.3 ha), critical premining slopes are 14° for Boddington, 10° for Huntly and 6° for Willowdale. Landforms exceeding these conditions may need site-specific designs to mitigate gully erosion risk. The rate of cumulative erosion and gully development measured by erosion pins on selected hillslopes closely follows the trend of cumulative precipitation at least during the first three seasons of rehabilitation growth. After this time, most gullies reach stasis. Cumulative erosion of non-mined, natural slopes also closely follows cumulative precipitation but at much lower rates (about 30 times lower than gullied sites). A proposed model of site erosion potential versus contributing area suggests that sites with the biggest gullies are above a threshold separating low-and highstate erosion. The effect of fire and maturity on the stability of gullied, rehabilitated sites is unknown.