Essex Specialities. 3: Annual Sea-purslane 131 there was a general contraction in range, particularly 1990-1993, related in part to the invasion of the more open salt marsh by Sea Couch-grass Elytrigia atherica and Red Fescue Festuca, rubra. The potential risks to the population were all-too-apparent from the outset: a result of its small size and extent (stock grazing; ditch flooding and pollution; spray drift) and position (catastrophic sea wall failure), and we rapidly came to the conclusion that we must not have all our eggs in one basket. Plants growing at Cambridge and, later, seed in the seedbank at Wakehurst Place gave some security, but we also felt that a second site would be desirable. And we had the ideal place next door, Foulness SSSI, protected by very large fences and UXBs. What's more, there were some ideal looking lagoons. So permissions were obtained and the translocation took place, with immediate success (Gibson 1991). Until 1993, when we turned up to monitor the introduction, only to find that a well-meaning islander had bulldozed the site to improve it for wildfowl. That population terminated. Still, we could rely on the native site .. . until we were greeted by a sward of Sea Couch and only some 400 plants, many very small, the lowest count hitherto. An executive, 'on-site' decision was taken: a series of turves were removed by hand, and seed moved the 10m or so into these bare areas. Also that year, seed was taken under licence into cultivation by Eric Steer and bulked-up very successfully with a view to future translocations and restocking. Our brinkmanship worked, though more through luck than design. Within two years, plants on the native plots had disappeared completely, whilst the adjacent scrapes were performing well (Fig.l); as a result further local 'gardening' has taken place to try and sustain the native site. Eric proved to be a maestro at producing large numbers of seeds, so by 1994 we were already in a position to try once more to establish a second population. We again selected the Foulness 'safe' site, this time making sure both the MoD and the islanders were aware of it; this has proved successful, thriving to date, and giving opportunity for us to investigate various features of its population dynamics and ecology through links with University College, London. Figure 1: Results of annual counts of Atriplex pedunculata at the native site in south Essex original site adjacent plots And with the high output of cultivated seed, we have been able to carry out trial restocking of former localities. Walberswick was first, but appears not to be too successful: perhaps this is not surprising as we do not really know why it became extinct there in the first place. Other possibilities on historical sites are also under consideration, although land claim has rendered a number of these unsuitable (eg Skirbeck, near Boston, is now a steel yard). Furthermore, we have also carried out trial Essex Naturalist (New Series) 17 (2000)