Assessment of observed and perceived changes in ecosystems over time, with special reference to the Sylt-Rّmّ Bight, German Wadden Sea

Examples of state changes in three aquatic ecosystems (the Neuse River estuary, NC, USA, the Kromme River estuary (St. Francis Bay, South Africa), the St. Marks National Wildlife Refuge, FL, USA) over time have been quantitatively assessed using ecological network analysis (ENA). A suite of ENA indices were compared among the multiple states of these ecosystems to illustrate the usefulness of ENA for describing differences among system states. Quantitative network models were constructed for the Sylt-Rømø Bight ecosystem and for the mussel bed subsystem of the Bight, depicting standing stocks of the living and non-living components and flow of carbon [a surrogate for energy] between them. These models consist of 59 compartments, and were assessed by means of ENA protocols. The impact of invasive species, such as the Pacific oyster, Crassostrea gigas, on existing mussel beds [of Mytilus edulis], and on predators [mainly birds] dependent on mussels for energy were assessed using AUTOMOD, a routine that predict the impact of species on each other within the ecosystem model. Predictive modelling clearly shows variability in bird biomass due to change in abundance of their important prey species at lower trophic levels. The simulations illustrated that the numbers of eider duck and oystercatcher in the Bight could be reduced to 50% of their current abundance within about 10 years at a 20%–40% reduction in prey biomass. Ecosystem properties calculated using ENA for the initial and alternate state of the Bight and the mussel beds after a simulated reduction of 40% in prey biomass show a decline in virtually all metrics (Total System Throughput, Development Capacity, Ascendency, Redundancy), system trophic efficiency, cycling, and system ratios (e.g. Average Mutual Information, Flow Diversity, Food Web Connectance).