GEOGRAPHIC VARIATION IN POSITIVE AND NEGATIVE INTERACTIONS AMONG SALT MARSH PLANTS

A pressing problem for ecologists is determining whether our understanding of communities, often developed from work at fine scales, can explain processes across broader scales. Here, we explore whether models of positive interactions developed in southern New England can be applied to geographic scales. Salt marsh plants may interact positively by ameliorating harsh physical conditions such as salinity stress. Because marsh porewater salinities are higher in low‐ vs. high‐latitude marshes, we tested the hypothesis that positive interactions are increasingly important at low latitudes. Previous work supported this hypothesis at the regional scale: within New England, positive interactions were rare in Maine but common in Rhode Island. We conducted parallel experiments in Georgia and Alabama to determine whether similar results would be obtained in a geographic comparison. Nine species of salt marsh plants were transplanted into control plots or plots with neighboring vegetation removed, in three marsh zones, at two sites each in Georgia and Alabama. Removing neighbors increased porewater salinities; nevertheless, plants usually performed best in neighbor‐removal treatments, indicating that most interactions were competitive rather than facilitative. Several mechanisms might explain these results. For widely distributed species, southern individuals may be more salt tolerant than northern conspecifics. We tested this hypothesis by comparing the salt tolerance of northern and southern conspecifics of three species in a common garden experiment. In two species, southern plants were more salt tolerant than northern conspecifics. Although this pattern may explain the rarity of facilitation in the south for some species, it cannot explain our overall results because few species were studied in both geographic regions. The most likely explanation for our results is that northern marsh floras are dominated by salt‐sensitive species that are likely to be facilitated by neighbors, whereas southern marsh floras are dominated by salt‐tolerant species that are unlikely to benefit substantially from neighbor amelioration of soil salinities. These results illustrate the difficulties inherent in extrapolating results from even well understood systems to different geographic locations or scales. Intraspecific adaptations and community composition are often likely to vary geographically, complicating our efforts to construct geographically robust generalities about community structure and processes. Corresponding Editor: D. Ayres.