Plant community composition and structure under short-term grazing exclusion in steppic arid rangelands

Grazing exclusion has been proved to be an effective approach for naturally restoring degraded rangelands. Still, the effect of this management practice on plant community composition and structure is ambiguous, especially under prolonged and repeated drought events. Here, we investigated the responses of community composition (i.e. incidence-based) and structure (i.e. abundance-based) to short-term sheep grazing exclusion under severe drought episode in the arid steppes of Alfa-grass (Stipa tenacissima L.) with a long evolutionary history of grazing. Individual species responses were tested based on species occurrence and abundance in either grazed or grazing-excluded steppes. Besides, indicator species analysis was used to identify species indicative of grazing-excluded steppes. Likewise, incidence-based and abundance-based alpha-diversity, beta-diversity and functional groups' diversities were quantified using Hill Numbers and compared between the two steppe managements types. Under severe drought conditions, sheep grazing exclusion allowed the apparition of a large number of increasers, colonizers, and native indicator plant species. It also improved the size of regional species pool and increased overall incidence-based and abundance-based alpha-diversity. Moreover, grazing exclusion decreased the abundance-based beta-diversity at local scale but increased it at landscape scale. The incidence-based beta-diversity significantly decreased at the landscape scale. Grazing exclusion enabled a significant spatial structuration of abundance-based beta-diversity components by maintaining high balanced variation in species abundance at large spatial scale and greater abundance-gradient at fine-scale. Our results suggest that the implementation of shortterm grazing exclusion in arid steppes would be the appropriate management practice for vegetation and habitat during prolonged droughts since it permits the recovery of native plant species and affects positively the size of the regional species pool, the overall incidenceand abundance-based alpha-diversity as well as the abundance-based beta-diversity (chiefly at landscape scale).