Vegetation composition of crane-disturbed vs. intact patches
We found that biopedturbation by foraging cranes created habitat patches with different vegetation structure and species composition compared to undisturbed stands of dry alkaline grasslands. The disturbed patches were characterised by sparse vegetation with small cover of vascular plants and cryptogams. The cover of perennial graminoids decreased, while that of perennial forbs and short-lived graminoids and forbs increased as a result of crane ploughing. The availability of open microsites and the low cover of competitor perennial graminoid species probably supported the germination of several plant species which resulted in larger plant diversity in disturbed areas compared to intact grassland (Deák et al. 2011). Similar patterns were found in Peruvian deserts where avian biopedturbations by burrow-nesting birds created more favourable areas for seedling establishment compared to undisturbed patches (Rengifo-Faiffer & Arana 2019).
In their meta-analysis Romero et al. (2014) showed that the overall effect of ecosystem engineers on diversity is positive and corresponds to a 25% increase in species richness, indicating that ecosystem engineering is a facilitative process globally. We detected a significant, approximately two-fold increase in plant diversity in the crane-disturbed plots compared to intact grasslands. This is particularly interesting, as such strong effect of engineer species has not been found for foraging bird species before and engineer effect was reported to be weaker at latitudes higher than 23º (Romero et al. 2014).
The species composition of disturbed and intact areas was clearly separated as shown by the multivariate analysis which implies that cranes create distinct habitat patches within the grasslands. Disturbed patches were characterised by early successional short-lived graminoids and forbs that are natural elements of the open, trampled and grazed patches of alkaline grasslands (Deák et al. 2015). Invasive or strong competitor weed species were absent on the disturbed plots, which might be due to the high salt content of the soil (see also Valkó et al. 2017). The disturbed plots were characterised by high inter-annual vegetation fluctuation, while this seasonality was less apparent in the intact grasslands.