Herbivore impacts on marsh production depend upon a compensatory continuum mediated by salinity stress

Plant communities are disturbed by several stressors and they are expected to be further impacted by increasing anthropogenic stress. The consequences of these stressors will depend, in part, upon the ability of plants to compensate for herbivory. Previous studies found that herbivore impacts on plants can vary from negative to positive because of environmental control of plant compensatory responses, a.k.a. the Compensatory Continuum Hypothesis. While these influential studies enhanced our appreciation of the dynamic nature of plant-herbivore interactions, they largely focused on the impact of resource limitation. This bias limits our ability to predict how other environmental factors will shape the impact of herbivory. We examined the role of salinity stress on herbivory of salt marsh cordgrass, Spartina foliosa, by an herbivore previously hypothesized to influence the success of restoration projects (the scale insect, Haliaspis spartinae). Using a combination of field and mesocosm manipulations of scales and salinity, we measured how these factors affected Spartina growth and timing of senescence. In mesocosm studies, Spartina overcompensated for herbivory by growing taller shoots at low salinities but the impact of scales on plants switched from positive to neutral with increasing salinity stress. In field studies of intermediate salinities, scales reduced Spartina growth and increased the rate of senescence. Experimental salinity additions at this field site returned the impact of scales to neutral. Because salinity decreased scale densities, the switch in impact of scales on Spartina with increasing salinity was not simply a linear function of scale abundance. Thus, the impact of scales on primary production depended strongly upon environmental context because intermediate salinity stress prevented plant compensatory responses to herbivory. Understanding this context-dependency will be required if we are going to successfully predict the success of restoration efforts and the ecological consequences of anthropogenic disturbances.