The Reintroduction of Large Wood as a River Restoration Approach: Associations Between Large Wood, Hydraulics and Fish

Abstract

<p>Restoring river ecosystems degraded by human activities is big business, worth billions of dollars annually. Investment in river restoration has grown rapidly over the last 40 years, with increasing recognition of the value of river ecosystems and in particular, for the environmental, economic and social goods and services they provide. Despite the increase in investment in restoring rivers our understanding of what contributes to the ‘success’ or ‘failure’ of river restoration remains limited. There are few guidelines that outline how to measure the success of restoration projects or to set ‘success’ criteria. Process-based restoration activities are increasingly advocated in order to improve our understanding of the success or failure of river restoration approaches. </p> <p>This thesis proposes an interdisciplinary systems approach to improve our understanding of the effectiveness of river restoration strategies. Such an approach focuses on processes and linkages between different components that comprise river ecosystems. The research undertaken in this thesis applies an interdisciplinary systems approach to investigate the reintroduction of large wood into the bankfull channel of the Barwon-Darling River, Australia, as part of a river restoration strategy. To do this, the importance of understanding the natural physical character and distribution of large wood within the channel of the Barwon-Darling is outlined; next the influence of wood character and its distribution on reach scale hydraulic processes over a range of temporal scales is demonstrated; and, finally the response of fish communities to the reintroduction of large wood is assessed. The approach taken links the physical character of wood with hydraulic processes and an ecological response at the appropriate reach scale.</p> <p>The aim of this thesis was to assess the reintroduction of large wood on the hydraulic landscape and the response of fish communities in river reaches of the Barwon-Darling River, SE Australia. Results of the research did not detect an effect to the reintroduction of large wood into this low energy dryland river system in terms of hydraulic landscapes or fish communities. The study design and techniques employed in this thesis allows hypotheses as to why these non-effects occurred and how future restoration strategies using the reintroduction of large wood can be advanced. </p> <p>There are four research foci to this thesis. First, the physical character and pattern of large wood along a 210 km reach of the Barwon-Darling River is examined. This component of the thesis argues that the natural pattern and physical character of large wood along the study reach differs to that proposed in current river wood models; models based solely on high energy river systems from the Northern America and Europe. For this research component of the thesis, 7142 individual pieces of large wood along the study reach were identified and their physical character and location within the bankfull channel recorded. Results demonstrate the pattern of large wood in the Barwon-Darling River do not support current river wood models for high energy river systems; river systems where wood pieces are actively transported downstream. It is hypothesised from my research that wood in the Barwon-Darling River passively realign themselves to the direction of flow and this alignment is controlled by wood size, wood density, its height in the channel and river channel geomorphology. Wood pieces may shift vertically within the channel and may shift orientation. Thus, models that describe wood patterns and character in high energy river systems cannot be applied to low energy river systems as these systems act differently. </p> <p>Second, the influence of reintroducing large wood on reach scale hydraulic character over multiple time periods was determined. The reach scale hydraulic character, termed ‘hydraulic landscapes’ of 10 replicated reference (naturally wooded), control (unwooded) and managed (wood reintroduced) reaches were determined during three low flow periods. These periods were: prior to the reintroduction of large wood to managed reaches; several months after the reintroduction of large wood into the managed reaches; and, then more than four years after wood reintroduction and following several large flood events. Results show no significant difference in the character of hydraulic landscapes between the three time periods for each reach type (reference, control and managed) and, managed reaches were more similar in character to the control reaches. It was found that the pattern and character of large wood differs between the replicated reference (wooded), control (un-wooded) and managed (wood reintroduced) reaches. Because of these differences in the pattern and character of wood between the reaches the reintroduction of large wood did not significantly alter the low flow hydraulic landscape over time.</p> <p>Third, the influence of large wood on reach scale hydraulic landscape at ten different discharges was examined. Results show the reintroduction of large wood did not significantly influence the hydraulic landscape of the managed reaches. The hydraulic landscape of the managed reaches remained more similar to the control reaches than the reference reaches and this was the case across all discharges. The character of the hydraulic landscapes did change with discharge. Notable stepwise changes in the character of the hydraulic landscape did occur in response to increasing discharge in all reaches. The lack of an influence of reintroduced large wood on the character of the hydraulic landscapes at any discharge is suggested to be a result of the reintroduced wood not matching the natural character and spatial pattern of wood in the river channel of the Barwon-Darling River. These results highlight the importance of effectively mimicking natural patterns in the distribution of large wood in restoration attempts.</p> <p>Fourth, the thesis focuses on the ecological response to the reintroduction of large wood. In doing so, it contributes to our understanding of the influence of reintroduced large wood on fish communities over time in a low energy river system and the importance of using process-based approaches to understand the links between large wood, the hydraulic landscape and fish communities. Fish communities were sampled in all replicate reaches i.e., reference (wooded), control (un-wooded) and managed (wood reintroduced) reaches. Sampling was conducted prior to the reintroduction of large wood, several months following the reintroduction and then again after several large flood events. Results from these different time periods demonstrate the reintroduction of large wood had a limited influence on fish communities. There were no significant differences between reaches in the total abundance and species composition of fish, but there was for fish length. However, significant differences were detected in total abundance, species composition and fish length between the three sampling periods. No significant interactions were recorded between reaches and sampling periods for total abundance and species composition but there was for fish length. It is hypothesised that the lack of a response by fish communities was because the physical character and position of the reintroduced large wood did not effectively replicate that of reference reaches and this had no influence on providing functional hydraulic habitat. In addition, the high variability in fish assemblages through time, a likely response to hydrological variation, reduced the power of the study to detect differences between fish communities over the relatively short time period of this study.</p> <p>The findings of this thesis highlight the potential lasting effects of the removal of large wood from river systems and that this may not be easily rectified through inappropriate restoration. In addition, the interdisciplinary systems approach taken in this thesis allows not only the integration of the disciplines of hydrology, fluvial geomorphology and ecology in order to resolve issues of river restoration but also promotes the importance of understanding processes and linkages within river ecosystems at a range of scales.</p>