Vulnerability to eutrophication of a semi-annual life history: A lesson learnt from an extinct eelgrass (Zostera marina) population

A semi-annual eelgrass (Zostera marina L.) population became extinct in 2004. It had flourished for many decades at Terschelling in the western Wadden Sea, one of the most eutrophied locations where seagrass growth has been recorded. Semi-annual populations survive the winter season by seed (annual), and by incidental plant survival (semi-annual). We compared seed bank dynamics and fate of plants between this impacted site and a reference site in the winter of 1990-1991. Seed bank density at Terschelling was extremely low (5-35 seeds m⁻²) in comparison to the reference site (>60 seeds m⁻²) and also in comparison to seed bank densities of (semi-)annual eelgrass populations in other parts of the world. Plant survival during winter was nil. Nevertheless, the population more than doubled its area in 1991, implying maximum germination and seedling survival rates. However, from 1992 onwards the decline set in and continued - while the nutrient levels decreased. To establish the cause of the low seed bank density, we conducted a transplantation experiment in 2004 to study the relationship between seed production and macro-algal cover. The transplantation experiment showed a negative relationship between the survival of seed producing shoots and suffocation by macro-algae, which is associated with light limitation and unfavourable biogeochemical conditions. The plants died before they had started to produce seeds. Thus, it is likely that macro-algal cover was responsible for the low seed bank density found in Terschelling in 1990-1991. Both the recorded low seed bank density and absence of incidental plant survival during winter were related to eutrophication. These parameters must have been a severe bottleneck in the life history of the extinct population at the impacted site, particularly as Z. marina seed banks are transient. Therefore we deduce that this population had survived at the edge of collapse, and became extinct after a small, haphazard environmental change. We argue that its resilience during these years must have been due to (i) maximum germination and seedling survival rates and (ii) spatial spreading of risks: parts of the population may have survived at locally macro-algae-free spots from where the area could be recolonised. As a consequence, the timing of the collapse was unpredictable and did not synchronise with the eutrophication process. The lesson learnt for conservation is to recognise that eutrophication may be a cause for seagrass population collapse and its eventual extinction, even years after nutrient levels stabilised, or even decreased.