Graduation of David van Leeuwen

The Connectivity Framework as a Tool to plan Nature Restoration Measures: A graph-theory approach to assess aquatic habitat connectivity of the Sliedrechtse Biesbosch

• Professor of graduation: Dr. ir. B.C. van Prooijen

• Supervisors: Dr. ir E. Mosselman (TU Delft, Deltares), Dr. ir. C.J. Sloff (TU Delft, Deltares), Dr. ir. L.M. Stancanelli (TU Delft), Ir. A.C. Briele (Iv-Infra)

This study assesses the application of graph theory to examine the connectivity of aquatic habitat in the Sliedrechtse Biesbosch and preserve or improve the area’s ecological value. The study addresses the relation between hydrodynamics and ecology, and evaluates different definitions of connectivity. Graph theory provides a novel and promising approach to assess connectivity in aquatic environments. With increasing availability of data, more accurate and informative analyses can be carried out, resulting in more effective designs of restoration measures. The application of graph theory in such research has, as one of its main strengths, its visual accessibility to laypersons or others without expertise in interpreting numerical model results. Furthermore, graph theory can offer both a holistic system view of a water network by evaluating the system as a whole, as well as a local view which is possible by looking at the single nodes and edges in the graph. This enables investigation of the role of individual channels within the system and the identification of vulnerable spots within the network, limiting the availability of aquatic habitat.

Graph theory is applied in various fields of research. As numerous metrics exist to examine the properties of a graph, an introduction to the most important metrics is given in this study. Adequacy of metrics is dependent on the questions posed and parameters relevant for the specific topic to be investigated. It is shown that in aquatic habitat connectivity, metrics such as betweenness centrality and bridges are indicative to obtain a general view of the network. When temporal variations play a role, as is the case in a tidal area, metrics as the number of components (NOC), the order of the largest component and the length of connected pathways (LOCOP) of the largest component are suitable to determine the connectivity. Whereas these metrics show useful in studying aquatic habitat connectivity, they may be less appropriate in connectivity studies of the same water system aiming at other fields of application, e.g. sediment connectivity.

Results show that graph theory provides a useful instrument in analyzing the aquatic habitat connectivity of the Sliedrechtse Biesbosch, which is investigated in a case study. The ease at which key nodes and edges are identified offer great possibilities for the design of nature restoration measures. In the present layout of the study area, large variation of aquatic habitat connectivity occurs based on a flow velocity fragmentation threshold of 0.3 m/s, corresponding to the maximum tolerable flow velocity for the European flounder (Platichthys flesus). Due to tidal influences in the study area, flow velocities vary continuously and the threshold flow velocity is exceeded during part of the tidal cycle. Considering the available habitat of other species gives different results depending on the tolerable flow velocities of the specific species. As is shown in this research, a combination of graph theory and numerical modelling enables the design and simulation of different nature restoration measures and system layouts to improve the aquatic habitat connectivity of the area.

The method presented in this research can be particularly useful to ecologists investigating suitable habitats for specific fish species. Also for engineers and others involved in the design of nature restoration measures, the method can be helpful since the designs of restoration measures can be evaluated considering the effect on habitat availability. The research provides an informed basis for subsequent applications of graph theory and numerical modelling to aquatic habitat connectivity. By selecting the most suitable design parameters and improvements of the network schematization, a justified decision can be made on the most effective restoration measures concerning the improvement of available habitat. Especially considering a combination of habitat preferences, such as flow velocity, water depth and turbidity, can provide proper insight in the aquatic habitat connectivity of an area for specific species.