Entire-area spring burning versus abandonment in grasslands: butterfly responses associated with hibernating traits

Abstract

Entire-area spring burning is a common practice in many grasslands. Although there are many studies on the effects of entire-area spring burning on insects, the underlying mechanism of insect response is not well known. Herbivorous insect responses to spring burning are expected to be related to hibernating traits and grassland quality (host and nectar plants). I surveyed butterfly assemblages and vegetation characteristics (e.g., nectar flower occurrence, host plant richness, vegetation height, and vegetation coverage) as parameters of grassland quality for butterflies over 21 transects in 3 burned and 3 abandoned grasslands. Butterfly species were classified based on their hibernating traits, with a focus on endangered and grassland species. Although grassland quality as measured by nectar resources and host plants did not differ between the abandoned and burned grasslands, butterfly richness and abundance were lower in the burned grasslands. Each grassland, particularly two burned grasslands, supported a lower number of Red List grassland species, considering the presence of their host plants. Butterfly richness and abundance exhibited a unimodal response to vegetation height and were lower in burned grasslands; however, they were not related to other variables. Species associated with burned grasslands hibernated belowground or on the surface in sparse vegetation, whereas species associated with abandoned grasslands hibernated on the surface or aboveground. Therefore, although burning is an effective method to halt grassland decline, entire-area spring burning is detrimental to grassland butterfly species. Rotational patch fire and grassland restoration would be required to maintain grassland butterfly diversity.