Retrofitting biodiversity: Ecological engineering for management of urbanised systems

Abstract

The replacement of natural coastal habitats by seawalls is a significant and developing issue worldwide. In Sydney Harbour, more than 50% of the foreshore is seawalls, built to protect urban assets against storms and erosion. Seawalls support considerably fewer species, notably mobile invertebrates, in comparison to natural rocky shores. A key mechanism for the difference in assemblages between artificial and natural habitat is the heterogeneity of the substratum. Seawalls in particular are often vertical, flat, featureless structures, in direct contrast to natural shores, which have a variety of microhabitats such as crevices, rock pools and boulders. A global effort to mitigate negative ecological impacts of artificial structures has included ‘ecological or eco-engineering’, which is the multifunctional design of marine infrastructure to benefit both humans and nature. Eco-engineering, as applied to seawalls, aims to challenge the traditional homogeneous design of such infrastructure by incorporating habitats to enhance marine biodiversity. The aim of this study was to quantify the ecological benefits of engineering missing intertidal habitats into seawalls, and at what scale any benefits become detectable. Precast ‘flowerpots’ that retain water over the tidal cycle were retrofitted onto seawalls; theseenhancements were designed to incorporate artificial rock-pools into seawalls, thus increasing the range of habitat types present. Previous trials showed that the flowerpots significantly added biodiversity to the seawall.The work presented in this thesis advances knowledge by testing the benefits of eco-engineering at a spatial scale beyond the structure itself by considering the seawall as an entire unit and by quantifying the effect of adding artificial habitats to seawalls on fish assemblages. This latter consideration has received little attention to date. Additionally, the effectiveness of the flowerpots to conserve natural intertidal habitats was evaluated. Specifically the models tested were: 1. Species would colonise flowerpots that are not found on the seawall; 2. Features of the flowerpots would affect the temporal variation of the seawall assemblage; 3. Benthic assemblages would differ between artificial and natural rock pools; 4. Species found in natural rock pools, but not flowerpots would not recruit to flowerpots; and 5. Flowerpots would have an effect on the fish assemblage at a larger (between seawalls) and smaller (within a seawall) -spatial scale. Asymmetrical Beyond BACI experimental designs were used to test larger-scale effects of the flowerpots on benthic and fish assemblages. Flowerpots supported a significantly different benthic assemblage to that on seawalls. This may be a localised effect, as no consistent differences were found in the temporal variation of the seawall assemblage with flowerpots installed in comparison to control seawalls. The species in flowerpots were characteristic of those that would be found in rock pools on natural rocky shores. In a direct comparison between natural and artificial rock pools, however, assemblages were significantly different. In particular, the predominant cover in artificial rock pools was turf algae, whereas in natural rock pools it was rock, assumed to be covered in biofilm. Additionally, artificial rock pools had a significantly smaller abundance and number of species per sample of mobile herbivores. The mobile species not found in the flowerpots, were also not found as juveniles recruiting to artificial turfs deployed over one year in the flowerpots. This result highlights the importance of considering recruitment to eco-engineered structures, especially if the objective of the habitat-feature is to target a certain species assemblage. Fish at high tide visited flowerpots, and the results showed seawall and pot-scale effects of the flowerpots on the number of species and abundance of fish. The temporal variation in the abundance of fish at the seawall with flowerpots installed was significantly different to control seawalls. Further, smaller-scale differences were found in the abundance of some fish species at flowerpots in comparison to adjacent control sections of the same seawall. This result highlights the importance of considering spatial and temporal scales in eco-engineering research. Installed habitats can have an effect on biodiversity at the scale of the added habitat, or at larger scales. Typically, research to date has been limited to the scale of the added habitat; however, multiple spatial and temporal scales should be included in experimental designs to fully understand the impact of habitat manipulation. Further, information on the scales of complexity particular organisms respond to will maximise the efficacy of eco-engineering designs. The results presented provide evidence that artificial habitats added to seawalls can be used to enhance a wide variety of taxa. More research is needed however if the goals of eco-engineering are to conserve intertidal habitats, as the results suggest that the assemblages in artificial and natural habitats have important differences. Clear management objectives are needed in eco-engineering research, and this thesis has aimed to inform end-users about using habitat enhancements as one tool to sustain marine biodiversity in urban systems.