Ecosystems effects 25 years after Chernobyl: pollinators, fruit set and recruitment.

Animals are assumed to play a key role in ecosystem functioning through their effects on seed set, seed consumption, seed dispersal, and maintenance of plant communities. However, there are no studies investigating the consequences of animal scarcity on seed set, seed consumption and seed dispersal at large geographical scales. We exploited the unprecedented scarcity of pollinating bumblebees and butterflies in the vicinity of Chernobyl, Ukraine, linked to the effects of radiation on pollinator abundance, to test for effects of pollinator abundance on the ecosystem. There were considerably fewer pollinating insects in areas with high levels of radiation. Fruit trees and bushes (apple Malus domestica, pear Pyrus communis, rowan Sorbus aucuparia, wild rose Rosa rugosa, twistingwood Viburnum lantana, and European cranberry bush Viburnum opulus) that are all pollinated by insects produced fewer fruit in highly radioactively contaminated areas, partly linked to the local reduction in abundance of pollinators. This was the case even when controlling for the fact that fruit trees were generally smaller in more contaminated areas. Fruit-eating birds like thrushes and warblers that are known seed dispersers were less numerous in areas with lower fruit abundance, even after controlling for the effects of radiation, providing a direct link between radiation, pollinator abundance, fruit abundance and abundance of frugivores. Given that the Chernobyl disaster happened 25 years ago, one would predict reduced local recruitment of fruit trees if fruit set has been persistently depressed during that period; indeed, local recruitment was negatively related to the level of radiation and positively to the local level of fruit set. The patterns at the level of trees were replicated at the level of villages across the study site. This study provides the first large-scale study of the effects of a suppressed pollinator community on ecosystem functioning.