An intermediate level of disturbance (ID) is considered to promote the greatest biodiversity in communities, as both competitive and colonizing species can coexist. This hypothesis (IDH) has been controversially discussed recently, in part because living disturbance agents are often also consumers, and the effects of disturbance are difficult to disentangle from consumptive effects. We studied the impact of a graminivorous (consuming) and semi-fossorial (disturbing) small rodent species, the common vole Microtus arvalis, over a 5-year succession using a replicated experiment with 16 enclosed, sown grassland plots of 2.4 m² (96 plant surveys). Eight plots received annual, short-term (13 ± 12 days) vole treatments in autumn, mimicking high autumn densities of fluctuating vole populations (disturbed), while the other 8 plots remained undisturbed. We studied a) combined consumptive and disturbance effects in different functional plant groups, vegetation structure, and on fodder value to estimate the provisioning service of the grassland. Disturbance affected succession processes, shifting the vegetation composition, decreasing vegetation cover, and facilitating persistence of rare, flowering, and ruderal plants, which had declined over time in undisturbed plots. While local plant diversity (α-diversity, Shannon index) was not affected, γ-diversity of joint disturbed plots and differentiation among disturbed plots (β-diversity) was doubled, compared to the joint undisturbed plots. Vole activity decreased agricultural fodder value of the grassland. Individual voles varied in digging activity, mobilizing between 0 and 4.5 L of soil while on the plot. We quantified the volumes of disturbed soil. Intermediate disturbance levels in autumn decreased the annual turnover rate of plant species during succession and promoted higher plant richness in the following summer, while both low and high disturbance levels reduced diversity and increased turnover. Through the burrowing and selective foraging by voles, disturbance facilitated a higher plant diversity by maintaining rare and flowering plants, disrupting the successional pattern. The dataset consists of two parts. The first part covers all 16 plots and presents the vegetation data for each plot for each of the five observation years. The second part refers only to the eight plots that hosted a vole for a short time in each observation year and presents variables related to the voles’ burrowing activity.

Changes after Jun 11, 2026: The factor indicating whether a plot hosted a vole once a year or not was not specified correctly for the years 19 and 20. In the previous version there were no entries in the column vole_treatment for these years. This has been corrected so that the factor vole_treatment now properly indicates for each year and plot whether a vole was put into it once a year or not.