Concern about abrupt and potentially irreversible ecosystem thresholds and tipping points is increasing, as they may have significant implications for natural capital and human wellbeing. Although well established in theory, there are few empirical studies that provide evidence for these phenomena in coastal and estuarine ecosystems, despite their high value for provision of ecosystem services. To determine the likelihood of such events, we tested two statistical methods; sequential Ttest analysis (STARS) and generalized additive models (GAMs) in a harbour ecosystem. These methods were applied to time series data spanning up to 25 years coupled with analysis of the relationships between drivers and natural capital asset flows. Results of the STARS analysis identified nonlinear thresholds in three of the natural capital assets of the harbour; mudflat area, Manila clam stocks and wader/wildfowl numbers, as well as an increase in several drivers affecting the harbour. The most prominent threshold was recorded in the Manila clam fisheries of the harbour, with stocks in two locations of the the harbour declining by 73-78% between 2006-2008. We suggest that the historic decline in the Manila clam stocks of the harbour were partly attributable to illegal fishing pressure although other factors such as disease and lease bed holders switching to other species were also likely to have contributed. More recently (2015-onwards) wild clam stocks of the harbour have increased thanks to improved management measures by local authorities. Generalized additive models also identified the contribution of macroalgal mats, sediment shoaling and river flows to historic changes in mudflat area, saltmarsh area and wader/wildfowl numbers. We conclude that information on thresholds and tipping points obtained using these approaches can potentially be of value in a management context, by focusing attention on the interactions and positive feedbacks between drivers that may cause abrupt change in coastal ecosystems. 1 Introduction Concern about abrupt and potentially irreversible ecosystem transitions is growing rapidly, as they may have significant implications for human wellbeing and are forecast to increase with intensifying climatic change and environmental degradation (Scheffer et al., 2001; RockstrÖm et al., 2009). Such transitions may result from an abrupt change in underlying drivers (e.g. land cover change, nutrient inputs), from an interaction between drivers, or from an abrupt change in the state of the ecosystem with a small or smooth change in drivers (Andersen et al., 2009). Another possibility is a threshold driven by a positive feedback loop, which is often referred to as a tipping point (Scheffer et al., 2009; 2012). While identifying such thresholds and tipping points can be challenging to identify in practice, evidence is increasingly indicating that nonlinear threshold responses could be widespread. Incorporating information about such responses into management plans can facilitate improved management outcomes (Huggett, 2...