Management Options
Inappropriate discharge prevention
The most obvious method to reduce sediment inputs to the river is to eliminate sediment-laden stormwater discharges that currently enter the river. In particular, discharges should be targeted that discharge into the river upstream of, or in, spawning areas. Discharges that increase bed agradation in the middle reaches of the river should also be reduced if at all possible. On-site measures to reduce sediment loading in discharges to match background river levels should be investigated.
Achieving this may be simply a function of increased monitoring and stricter enforcement of existing discharge consents. While Couling & McCallum (1994) state that runoff from urban subdivisions requires strict on-site control, it was clear that current development projects are failing to meet the conditions of their discharge consents. During every significant rainfall event that occurred during the field component of this study, discharge turbidities of up to 508 NTU (suspended solids c. 295 mgL-1) were measured in discharges from the old Applefield’s orchard subdivision (Plates 8 & 9).
The extensive box-drain network that takes stormwater from urban areas to the Styx could be redeveloped to include settling ponds and a more natural, sinuous channel configuration. This would reduce water velocities and cause suspended sediment to deposit in the drains instead of in the river itself. It is likely that occasional dredging in the settling ponds would be necessary to maintain their efficiency. Any attempt to make drains follow a sinuous configuration should take careful note of the terrestrial morphology so that artificial curves agree well with the terrain and will not result in significant bank erosion where water flows attempt to short-cut the artificial channel. Use of rip-rap or large woody debris may be necessary to protect banks on newly created meanders.
Plate 8. A discharge-affected tributary at its confluence with the main stem of the Styx River, 100m above Site B.
Plate 9. A mud-like discharge flowing into the main stem of the Styx River near Main North Road.
Riparian management
The Styx River was historically surrounded by extensive wetlands and forest. Sediment yields from wetlands and native forest tend to be much lower than other landuses (Scarsbrook & Halliday, 1999). Agricultural areas, for example, are known to be significant contributors of sediment to streams with attendant effects on stream ecology (e.g., Boulton et al., 1997; Scarsbrook & Halliday, 1999). Hicks and Duncan (1993) found that mature urban areas are even worse, contributing 1.5 times the sediment yield than pasture. While exact figures are not available, it is well accepted that sediment yields from construction sites, including urban subdivisions, are many times the yield from mature developments. Clearly, sediment inputs to the Styx River from surrounding land-use changes, have increased dramatically since the settlement of Christchurch. For example, plates 10 & 11 show the extent of riverbank erosion resulting from unrestricted cattle access to the river along the middle reach used in this study. This situation is common wherever stock have access to the river.
Recognition of landuse impacts on sedimentation in streams has increasingly led to the practice of planting riparian buffer zones, or corridors, as a tool for stream restoration (Williams et al., 1997). A corridor of trees and plants can absorb nutrients from subterranean water flow into the river, as well as acting to filter sediment out of any overland flow (runoff). Typically, such corridors need to extend out to approximately 10-15 meters, on both sides of the river, in order to achieve significant reductions in sediment and nutrient inputs (Cooper et al., 1987; Williams et al., 1997). Implementation of this concept would have multiple benefits to the ecology of the river and the surrounding area, as well as reducing non-point source inputs of sediment. A first step in this direction would be the construction of fences to exclude cattle from the river.
When planning riparian corridors it should be remembered that the width of the corridors must be increased to compensate for steep terrain. For example, riparian buffer zones for forestry activities in the states of Oregon and Washington.
Plate 10. Some bank erosion caused by cattle, upstream of Site B.
Plate 11. More cattle erosion upstream from Site B.
Washington extend up to 200m from river edges (Williams et al., 1997). Riparian corridors are very successful at reducing sediment inputs in the long-term (Cooper et al., 1987), but river widening has been noticed in some studies and may potentially increase bank erosion in the short-term. Also, the benefits associated with riparian corridors diminish rapidly if the corridor is fragmented or discontinuous along the river continuum (Scarsbrook & Halliday, 1999).
A further reason for establishing tall riparian vegetation is that it provides shading of the river bed. While not previously considered in this study, dense aquatic weed growths can contribute to increased sediment deposition by trapping suspended sediment. There are presently extensive growths of aquatic weed in the middle reach and these are almost certainly contributing to the embeddedness of the river gravels. It is unknown if these dense weed growths have become worse over the past decade. Shading the riverbed may help to reduce the extent and density of the weed beds, and reduce sediment trapping.
Instream habitat improvement
Trout and salmon prefer to spawn in reaches which alternate between pools and riffles (Alonso et al., 1996; Lisle, 1989) and often construct redds in the tailspills of shallow pools to take advantage of the flow characteristics associated with this situation. For instance, reduced water velocities in the pools allow heavier sediments to settle out of the water column before reaching the redd. Some habitat of this nature is present in the upstream reach of the study but the middle reach is a relatively unbroken, straight run of water (Eldon & Taylor, 1990) with limited habitat variation and shelter. A wider range of habitats is likely to encourage a wider diversity of biota in the river, and also provide refuges from predation and storm flows.
Instream habitat improvement through the addition of large woody debris (LWD), boulders etc., has been successfully used in watershed restoration projects overseas to increase habitat variability, and also to increase the extent of spawning habitat available to trout and salmon (Williams et al., 1997). Such measures may also be helpful in the Styx. For example, tree stumps, & logs etc. have been absent from the river since the surrounding native forests and wetlands were cleared. As well as creating more variable flow conditions (that enable trout to position redds in more favourable situations), the introduction of LWD may provide benefits for species such as the freshwater crayfish (Paranephrops zealandicus), and native fish.
River dredging
Sediment dredging in the middle and lower reaches of the Styx River has occurred on a sporadic basis during the last century (Hicks & Duncan, 1993), but no dredging activities have been recorded above Radcliffe Road. It is possible that removing the accumulated deep sediments between Radcliffe Road and Main North Road may serve to increase the hydraulic gradient and possibly shift sediment deposition back downstream. Clearly this is a drastic measure with associated effects on downstream turbidity and possibly effects on adult trout in the lower reaches of the river. It is also a short-term response and fails to address the problem of increased sedimentation in the Styx River. However, it may be helpful in restoring sediment deposition regimes in the short term.