Quantifying and predicting benthic enrichment : lessons learnt from southern temperate aquaculture systems

The increasing demand for aquaculture products globally is leading to greater demand for coastal marine farm space, intensification within existing aquaculture areas, and conversion of production to high value species, especially finfish. Among the many environmental interactions that arise with finfish aquaculture development, one of the most dramatic impacts is local-scale organic enrichment of the benthic ecosystem due to deposition of fish faeces and uneaten feed. A benthic impact is typically evident as severe organic enrichment beneath finfish cages (e.g. species-poor, near-azoic conditions), with a gradient of decreasing enrichment extending to background conditions across scales of tens to hundreds of metres distant from cages. The overall hypothesis of this thesis was that seabed organic enrichment (degradation and recovery) can be accurately and quantitatively determined using biological and physicochemical variables that can be applied across geographic regions and contrasting environments. This was accompanied by an objective to refine knowledge of processes underpinning benthic enrichment, and to develop or refine tools for the prediction, monitoring and management of enrichment effects associated with fish aquaculture. The thesis comprises six sequential, related chapters that address: site- and region-specific ecological characterisation of benthic communities and the development on a new environmental indicator variable; comparisons of existing biological indicators and indices in different hydrodynamic regimes; application and validation of a depositional model for predicting effects under very different environmental conditions; and a detailed analysis of long-term and medium-term recovery from highly enriched states, and consideration of re-impact rates and implications for farm management strategies. The analyses are based on both targeted recent studies as well as longer-term monitoring undertaken at six salmon farms situated in the Marlborough Sounds, New Zealand; four of which are situated in low flow environments, and two are situated in high flow (dispersive) environments. Characterising the differences associated with the sites' dispersive properties is a theme that runs throughout this study. Chapter 2 used best professional judgement methods to develop a quantitative benthic enrichment index termed 'enrichment stage', which unifies information from biological and physico-chemical variables. The resulting seven stage bounded continuous variable was used to assign enrichment tolerance groups to benthic taxa using quantile regression splines. A number of key indicator taxa were discriminated along the enrichment gradient, including several that were responsive to low-level changes in enrichment stage (ES), but not necessarily organic matter (%OM), and 10 taxa for which ecological understanding was previously limited. In Chapter 3, the gradient was also used to evaluate the performance of five benthic indicators and ten biotic indices for defining organic enrichment under different flow regimes. A subset of variables was recommended comprising: two biotic indices, total abundance, and a geochemical variable. A subsequent but related study in Chapter 4 revealed pronounced flowrelated differences in the magnitude and spatial extent of benthic enrichment. Total macrofaunal abundances at high-flow sites were nearly an order of magnitude greater than at comparable low flow sites, representing a significant benthic biomass, and occurred in conjunction with moderate-to-high species richness and the absence of appreciable organic accumulation. The atypical ecological conditions associated with high-flow sites were attributed to i) minimal accumulation of fine sediments, ii) maintenance of aerobic conditions in near-surface sediments, and iii) an abundant food supply. Chapter 5 explored the relationship between predicted depositional flux (using DEPOMOD) and enrichment stage, calculated using the methods developed in the previous chapters (1 to 3). Observed impacts at farms with contrasting flow regimes were examined to evaluate the role of modelled resuspension dynamics in determining impacts. When resuspension was included in the model, net particle export was predicted at the most dispersive sites. However, significant seabed effects were observed, suggesting that although the model outputs were theoretically plausible they were inconsistent with the observational data. When the model was run without resuspension the results were consistent with the field survey data. This retrospective validation suggested that approximately twice the flux was needed to induce an effect level at the dispersive sites equivalent to that at the nondispersive sites. Flux estimates are provided for detectable enrichment and highly enriched states. This study shows that the association between current flow, sediment resuspension and ecological impacts is more complex than presently encapsulated within DEPOMOD and emphasises the need for validation of such models, particularly at dispersive sites. The final two data chapters (Chapters 6 and 7) examine the spatial and temporal recovery processes that take place following a highly enriched state. Chapter 6 provides a comprehensive analysis of a long-term (8 year) dataset in relation to a variety of proposed recovery and remediation definitions. Many challenges associated with quantifying the endpoint of 'recovery' were identified. The concept of dynamic and spatial equilibria proved to be valid in this situation, and alternate state theories may apply. In combination with visualisation of plotted data, statistical tests for parallelism in temporal trajectories of cage and reference sites proved to be an effective method for characterising recovery, but the method was highly sensitive to window time-length. Simple, univariate indicators of enrichment tended to be less sensitive, and indicate recovery earlier, than more complex multivariate indicators. Recovery was assessed to be complete after approximately five years, but there was some evidence of on-going instability in the composition of the macrofauna, which was partly attributed to spatial and temporal processes and patterning in the macrobenthos. The last data chapter (Chapter 7) examined shorter-term recovery and re-impact patterns and revealed some interesting successional patterns in time and space, especially between %OM, TFS and abundances of opportunistic taxa. The discussion brings together findings from the targeted and long-term studies to reveal alternate oscillations between sediment chemistry and biological response, which have temporally distinct signals. It is proposed that the large oscillations that occur in the early stages of recovery represent the extreme end of the environmental instability that occurs as a result of a severe perturbation (in this case, cessation of extreme enrichment) that abates through time as recovery ensues. This integrated study has a number of important implications for the management of organic enrichment in general but is especially pertinent for fish farming. In particular, recommendations are made regarding the i) adequacy of chemical and biological benthic indicators and their performance in typical non-dispersive and atypical dispersive sites; ii) use and applicability of depositional models in the same environments with emphasis on the role of resuspension, and iii) timing and approach for reintroduction of impacts, with respect to monitoring and management of rotational fallowing strategies to ensure on-going sustainability.