Ecological disturbances are recognized as a crucial factor influencing the attributes of ecological communities. Depending on the specific adaptation or life cycle, plant species show different responses to disturbances of different magnitudes. Herben et al. (2016) proposed disturbance indicator values (DIVs) that describe the niche of each Central-European plant species along gradients of disturbance frequency and severity. Here we ask if the DIVs can be used in community ecology for bioindication of disturbance regime?

We used a dataset of riparian forests sampled within mountain catchments (the Sudetes, SW Poland). As the regime of disturbance is driven by changes in floods from the spring towards the mouth, we calculated the position of every plot along longitudinal (upstream-downstream) gradient and used it as a proxy for the disturbance severity and frequency. We then calculated the community-weighted means (CWMs) for each of the six indices for each plot and analyzed whether these indices reflected the position of the plots along the rivers. We expected an increase in the severity indices and a decrease in the frequency indices downstream along the rivers. Moreover, we analyzed relationships between disturbance indices and species optima along longitudinal gradient.

Surprisingly, means for all analyzed indices increased along the rivers. Severity indices showed the strongest association with the longitudinal gradient. The disturbance severity index for herbs was the only index that differed significantly among species with different responses along longitudinal gradient. On these results, we identified a strong correlation between the severity and frequency indices as the main problem.

We conclude that the DIVs have considerable applicative potential; however, the determination of ecological niches separately for disturbance severity and frequency is difficult because different components interact to shape the realized niche of each species. All analyzed indices encompass different attributes of the disturbance regime including both severity and frequency.

The riparian forests were surveyed along mountain rivers in the Polish part of the Sudetes, SW Poland. The dataset used in this study was collected between 2006 and 2009 following the methods of the Central-European phytosociology (Dzwonko, 2007; Kent, 2012). All plots were located close to the river bank; plots were rectangular (7.5 × 20 m) and oriented with the long dimension parallel to the upstream-downstream gradient of the river. The cover of vascular plants was estimated for each of three layers (trees, shrubs and herbs) using the Domin-Krajina scale with an ordinal transformation proposed by van der Maarel (1979). Plots were also located with a GPS receiver and recorded in a mobile GIS database. The precise localization and digital river network (Map of Hydrographic Division of Poland, MPHP) enabled the calculation of the distance from the river source for all the sampled plots. More details regarding study area and vegetation sampling are available in the previous study (Pielech et al., 2015). The full dataset is stored in the Forest Database of Southern Poland (Pielech et al., 2018). Values of DIVs are taken from Herben et al. (2016).

The extended dataset is stored in the Forest Database of Southern Poland (Pielech et al., 2018). It includes also spring forest and a broad array of environmental variables.