Question Does the effect of droppings and herbivory by wild guinea pigs on salt‐marsh vegetation vary across microhabitats? Location Upper salt marsh in the Mar Chiquita coastal lagoon (37°44′52″ S, 57°26′6″ W, Argentina). Methods During autumn 2014, we performed a micro‐scale experiment manipulating droppings of wild guinea pigs (Cavia aperea). After 15 days, we compared the content of nutrients (i.e., NH4+, NO3−, and PO43−) in soil. Then, we started two factorial experiments in the high marsh, specifically one in marsh matrix (i.e., tall vegetation) and the other in open patches (i.e., short vegetation). Units were randomly assigned to different treatments (with and without herbivores and droppings). In early summer, we compared total plant cover, maximum height, above‐ground biomass, plant species richness and the composition of plant assemblages between treatments. Results Droppings increased the content of PO43− in soil by five times, but decreased the content of NO3−. Herbivory reduced total cover, plant height, and above‐ground biomass in the marsh matrix and the open patches. In the marsh matrix, droppings increased the total cover, whereas herbivory reduced the abundance of Spartina densiflora (cover and above‐ground biomass) and subordinate species (cover). Additionally, the cover of subordinate species was negatively associated with the cover of Spartina densiflora. Therefore, in the marsh matrix, by increasing light availability at the ground level, herbivory increased plant species richness, but with low cover of subordinate species. In contrast, in open patches, herbivory and droppings interacted to reduce plant species richness. Conclusion Herbivory by small mammals can control the above‐ground biomass, structure, and plant community composition of the high marsh, but non‐trophic mechanisms (droppings) can also affect community composition. Nonetheless, ecological impacts of small mammals depend on the context, because the magnitude and direction of trophic (herbivory) and non‐trophic (droppings) effects change across microhabitats. Our understanding of trophic and non‐trophic effects of small herbivores remains relatively poor compared to that of large herbivores. Our experiment, simultaneously manipulating wild guinea pig herbivory and droppings in a salt marsh, shows that these factors can interact, and the magnitude and the direction of these trophic and non‐trophic ecological effects depend on the environmental context.