Predicting estuarine faunal assemblages using enduring environmental surrogates, with applications in systematic conservation planning

Conservation planning in estuaries has lagged behind that in terrestrial and marine areas, despite these valuable ecosystems being among the most degraded. The core of any such approach is a reliable habitat classification and inventory of target biota. These data, and particularly the latter, are often lacking at the local-regional scales most relevant to estuarine management. This study presents a quantitative approach for predicting the likely fish and benthic invertebrate assemblages at any unsampled estuarine site using readily-obtainable and enduring biophysical attributes. We apply this scheme to an urbanised estuary and predict the above faunas throughout its entire nearshore zone. These data are then used to systematically design an exploratory spatially-efficient reserve that meets representation targets for numerous faunal conservation features, and test the ability of an existing, unsystematically-derived Marine Park to do the same. Spatial patterns in the enduring biophysical attributes of local-scale habitats provided good to excellent surrogates for those in the fish and invertebrate faunas. All unsampled sites were then successfully assigned to their respective habitat and correlated fauna using biophysical measurements and a predictive decision tree. The resultant spatially-continuous ‘faunal map’ enabled quantification of 67 conservation features, from which reserves aimed at representing 10–30% of each feature were systematically derived. All reserves were highly efficient and almost always met representation targets, contrasting with the existing Marine Park which underrepresented 40–80% of features. Reserve designs were, however, spread throughout the system, highlighting the complexities in designing representative reserves for estuarine environments that capture their spatio-temporal diversity