Many studies have reported positive effects of agri-environment schemes (AES) on species richness and abundance, but studies demonstrating demographic, population-level responses, which is the desired outcome from AES, are rare. This is at least partly due to the difficulty in distinguishing between behavioural and demographic effects in snapshot studies. Here we report positive population-level effects of environmental set-asides (currently included in the national AES of fallows in Finland) in grassland butterflies and day-active moths.

We firstly identified a set of conditions that a species should meet in order to succeed in establishing a local population in an area where the focal agri-environment scheme has been applied. Secondly, we identified indicators for distinguishing the species that show true population growth from species merely visiting the AES area from the surrounding landscape. We then applied these indicators to two long-term butterfly and moth monitoring data sets (totaling 6925 butterfly and 11232 moth individuals) originating from (i) an eight-year set-aside experiment and (ii) a large, >10 years old set-aside field, thus identifying the grassland species that were capable of forming breeding local populations on long-term set-asides. We then tested predictions based on the two long-term monitoring data sets in another, independent butterfly and moth snapshot data set originating from 78 set-aside fields of different age.

Twenty-five out of the observed 65 species showed evidence of local population establishment, and 24 (96%) of these are considered as species preferring semi-natural grassland as their habitat. A large proportion of all the observed grassland-preferring species (53% of butterflies and 65% of moths) indicated local population establishment. As predicted, the probability of occupancy of the local population establishing species increased with set-aside age in the 78 set-asides. Moreover, the species establishing populations had a more strongly increasing occurrence probability with set-aside age than the species observed to occur on the experimental set-aside without actually breeding there.

Our results suggest that a substantial proportion of the Finnish grassland butterflies and moths form local populations in long-term environmental fallows of the Finnish national agri-environment scheme. Nine butterfly species with evidence of local population establishment on the set-asides comprise 25% of the 36 grassland-preferring species occurring in Finland. The average occupancy probability of these grassland species was 38% in the 45 out of 78 set-asides that were at least 5 years old. Our results indicate that the Finnish environmental set-aside scheme covering ca 5% of the national agricultural field area has positive population-level effects in many common grassland species. Thus it forms a significant means of supporting grassland biodiversity besides conservation management of semi-natural grasslands.