VEGETATION SUCCESSION AND HERBIVORY IN A SALT-MARSH - CHANGES INDUCEDBY SEA-LEVEL RISE AND SILT DEPOSITION ALONG AN ELEVATIONAL GRADIENT

1 The relationships between soil development, vertical vegetation zona tion, vegetation succession and herbivory by Brent geese, Branta berni cla, were studied in a coastal salt marsh. We were able to analyse up to 100 years of salt marsh development by comparing sites where vegeta tion succession had progressed for-various periods of time. These data were related to a continuous daily record of high water levels measur ed since 1824. 2 Most elevational variation in edaphic conditions (and therefore vertical vegetation zonation) could be attributed to variat ion in height of the sandy subsoil, as rapid dune formation occurs on the beaches early in succession. In the intermediate part of this elev ational gradient, the maximum annual increase of 1.2 mm of silt corres ponded to an annual increase of 5.6 g N m(-2) in the topsoil (0-50 cm) . The average sea level rise in this area over the last 170 years was 0.63 mm year(-1). A sedimentation model suggests that this has had str ong effects on sedimentation and the annual inundation frequency in th e mid-part of the elevational gradient, thus affecting vegetation zona tion on the salt marsh. For the major part of the investigated transec ts, sea level rise has probably speeded up succession due to an increa sed rate of sedimentation. 3 The occurrence and dominance of all plant species were recorded in 3927 plots, and and for the 11 most common s pecies response surfaces were calculated for their dependence on eleva tion and transect age. Most plant species were clearly separated along these axes. Most halophytic species, which were preferred by the gees e, occurred early in succession and low on the gradient, where we obse rved the highest densities ge quality of Festuca I rubra increased tow ards the lower salt marsh. Other preferred forage species (Puccinellia maritima and Plantago maritima) were gradually displaced during succe ssion by the tall grass Elymus athericus, especially in the mid-and up per salt marsh. Few geese grazed in areas where Elymus was dominant. 4 Herbivores first increased in numbers but then declined along a gradi ent of primary productivity. We propose that declining forage quality (due to changing vegetation composition during succession) is a better explanation for this pattern than the classic explanation of predator control of herbivores at high levels of primary productivity. This qu ality threshold hypothesis, as an alternative explanation of the explo itation ecosystem hypothesis, is expected to hold especially where sma ller (quality-sensitive) herbivores such as geese are present. 5 Grazi ng by cattle in a 200-year-old part of the salt marsh led to the disap pearance of Elymus athericus, to establishment of early successional h alophytes and a return of Brent geese. Crazing by a larger herbivore t herefore facilitated conditions for smaller herbivores by preventing t he dominance of plant species that were good light competitors, and th us improved forage quality. Populations of these small herbivores coul d then become regulated by predators, although none was present at our site.