Figure 5. Looking upstream at VRW2 (a low gradient rock weir) under low (top photo), intermediate (middle photo), and high (bottom photo) water level conditions. It is evident that orifice flow is the only active flow regime under low water level conditions, while orifice, gap, and over-weir flow are active simultaneously under intermediate and high-water level conditions. VRW2 under low water level conditions demonstrates the importance of embeddedness for enhancing fish passage effectiveness, while VRW2 under high water level conditions demonstrates the effect of ‘drowned conditions’.

Orifices

Orifices are challenging to survey in the field post-construction, and were assumed to be minor with regards to facilitating fish passage in the present study. In various river restoration projects, impermeable geotextile layers are installed at the stream bed and upstream of the weir crest to prevent orifice flow. However, this was not the case in Weslie Creek, where geotextile layers were not used in the rock weir design. Rather, smaller keystones and river stones were used to control flow. Although the design concepts for Weslie Creek indicate that keystones and footer stones are compacted with a mixture of 90 mm – 225 mm river stone, orifice flow is present based on field observations (i.e., VRW2). It is possible that the river stone placed in the orifices were transported downstream during large rain events, and consequently provided opportunities for orifice flow. Based on observations in the field, it is likely that orifices throughout the Weslie Creek rock weir system are < 0.05 m in width and depth. Literature recognizes rock weir designs that are purposely constructed with orifices as the preferred pathways for fish passage (e.g., Ead et al., 2004). Further, orifice design may enhance fish passage effectiveness via associated turbulent structures (Silva et al., 2012), or by providing space for energy dissipation through the structure.
Under low water level conditions, gap and over-weir flow may not be activated over or through rock weirs. During these times, orifice flow provides the only pathway for local fish species to travel upstream or downstream, depending on their life stage and behavioural characteristics. According to Kupferschmidt and Zhu (2017), velocity through orifices must be less than the maximum burst speed of the local fish species. This is also true for gap and over-weir flow. The difficulty associated with evaluating the effectiveness of orifices for fish passage, particularly during in-field analysis, is the inaccessibility for sampling equipment. Although orifice flow is considered non-negligible in Weslie Creek, sampling the required geometries and flow beneath keystones was not possible. Silva et al. (2012) analyzed fish passage effectiveness through orifices, however the research was conducted in a flume setting with the appropriate equipment for measuring velocity in such confined spaces.

Distance Between Rock Weirs

The characteristics of pool features surrounding rock weirs upstream and downstream are an important consideration for fish passability. According to Martens and Connolly (2010), suitable pool features provide refuge opportunities, habitat for rearing, and leaping pools for local fish species. The distance between rock weirs (i.e., the length of the pool feature) also influences flow through/over the downstream rock weir by dissipating energy and maximizing flow resistance (Wang et al., 2009). Pool features at Weslie Creek that were less than 4.0 m in length were more likely to produce flows that exceed local fish species’ burst swim speeds (m/s) and therefore reduce fish passability (Table 2). In contrast, the pool features that were greater than 4.0 m in length provide 100% fish passability under all water level conditions (Table 2). This is supported by literature that suggests pool length is the primary geometric dimension that influences flow through both conventional and nature-like fishways (Wang et al., 2009; Bermudez et al., 2010).
It was determined that as pool length increases, the total number of opportunities for fish habitat and/or refuge also increases, with few outlying instances. In terms of fish habitat and/or refuge, it is important to note that although recirculation zones were not identified in all pools under all water level conditions (Figure 3), the measured velocities were below fish species’ critical swim speeds. To recognize all possible locations for fish habitat and/or refuge in the Weslie Creek reach, further analysis is required to identify the sustained swim speeds for local fish species. For example, low cross-sectional velocity values (i.e., 0.02 m/s) were measured at pool features in Weslie Creek and most likely facilitate fish habitat and refuge, however the sustained swim speeds appropriate for local fish species are unknown. As such, only locations with stagnant or recirculation zones were used as indicators for fish habitat and/or refuge in Weslie Creek. Since sustained swim speeds are less than burst swim speeds, it is likely that such low cross-sectional velocity values do support local fish habitat and/or refuge conditions (Beamish, 1978).
In terms of pool length, there are conflicting goals between channel stability and fish passage (Thomas et al., 2000). With a greater pool length, the distance between rock weirs increases, and creates a greater drop height between rock weirs. Fewer rock weirs throughout the reach is problematic for channel stability due to a larger gradient. Additionally, fewer rock weirs throughout the reach is problematic for fish passage due to the greater drop heights local fish species are required to maneuver. It is recommended that pool lengths (the distance between rock weirs) be large enough to provide suitable conditions for passage, habitat, and refuge, but not undermine channel stability. The conclusion from this analysis should be applied in future natural channel design projects to ensure pool features are measured to an appropriate length to provide maximum opportunities for 100% fish passability and channel stability.

Evaluating Effective Rock Weir Design and Construction

According to Lucas and Baras (2008), river restoration efforts (such as fishways) should provide 90% overall passage efficiency for diadromous and potamodromous fish species to be considered functional. Fish passability through a reach is a function of three components: appropriate water depth, velocity, and gradient for leaping (Williams et al., 2012). It is important that such components are suitable for the target fish species within the system, both at the rock weir structures and pool features (Williams et al., 2012). The results of this research suggest that fish passability through the reach is most effective and longitudinal connectivity is most complete under low flow conditions (Figure 3). This is likely attributed to the number of active gaps available for fish passage. Although not all gaps facilitated the appropriate velocity for local fish passage, 9/10 rock weirs have at least one suitable pathway. With 90% overall passage efficiency based on local velocity measurements, the rock weirs at Weslie Creek are considered functional under low water level conditions. Excessive velocities through flow pathways that inhibit fish passage upstream is recognized as the most likely cause of passage failure and non-functionality through rock weir systems (Knaepkens et al., 2006). The mark-recapture of non-salmonid species through a pool-weir system in Belgium yielded 0%, 8%, and 29% fish passage effectiveness for bullhead (Cottus gobio ), perch (Perca fluviatilis ), and common roach (Rutilus rutilus ), respectively. These fish species are larger and have stronger swimming capabilities than species local to Weslie Creek. This demonstrates that despite the size of the fish species and their swimming and/or leaping capabilities, where velocities exceed burst swim speeds, the rock weir system is not functional.
The general consensus concerning rock weirs and fish passage is that there is a lack of standardized monitoring protocols for evaluating effectiveness or fish passability (Silva et al., 2018). Further, fish passage analyses are common in the literature, however their measure of effectiveness differs. The structural differences between rock weirs and other nature-like fishways also contributes to challenges for comparing fish passage results. For example, PIT-tagging is common for fish passage monitoring in larger species (e.g., Tummers at al 2016; Martens and Connolly, 2010). However, in small-bodied fish, such as those in Weslie Creek, a different approach is needed. Rather than observing where the fish go, the hydrodynamics and geometries of the system were used to evaluate fish passage feasibility given physiological abilities of the expected local fish populations. In addition to the different methods and the different measures of fish passage effectiveness, it is likely that rock weir structure and design differences contribute to differences in fish passability results. Target fish species with different burst swim speeds will require different conditions (i.e., rock weir design and water level) to facilitate fish passage. As such, a relationship may exist between the size/swimming characteristics of fish species and appropriate conditions for fish passage effectiveness. Large-bodied fish species require greater water depth and can employ stronger burst swim speeds to maneuver rock weirs, while small-bodied fish species require smaller water depth and require lower velocities through rock weirs for passage. Such relationships must be considered in rock weir designs for effective fish passage for the target fish community.