Control of the toxic plant Colchicum autumnale in semi-natural grasslands: effects of cutting treatments on demography and diversity.

Abstract

Semi-natural grasslands are important habitats for the conservation of biodiversity in Europe. High population densities of toxic Colchicum autumnale in these grasslands may cause problems for livestock and the marketing of hay. Consequently, farmers may either intensify grassland management to reduce C. autumnale in the fodder or abandon the land; both practices will lead to a loss of biodiversity. Previous studies suggesting early cutting to control C. autumnale did not consider population dynamics and the effects on plant diversity. We conducted a four-year experiment in six regions within Austria and Germany, applying five cutting treatments in 16 C. autumnale populations to test the effects of cutting date and frequency on C. autumnale and co-occurring vegetation. Demographic data were evaluated with matrix population models, life-table response experiment (LTRE), ANOVA and MANOVA. Vegetation data were analysed with multiresponse permutation procedures (MRPP), ANOVA and MANOVA. Population growth rate was significantly reduced in plots cut in early and late May compared to plots cut in June (control). Plants cut in late April or early May showed the lowest survival probability. Significantly fewer large vegetative plants developed capsules in the following year when cut in early or late May. LTRE analysis showed that differences in the population growth rate between the control and early cut treatments were mainly the result of a reduced survival and growth and an increased retrogression to smaller stages. Multiresponse permutation procedures revealed no differences in vegetation composition between treatments except for one site in 2011. There were no differences in Shannon index, evenness or species turnover rate within any year. Synthesis and applications. The greatest reduction in vitality of Colchicum autumnale was observed in grasslands cut in late April or early May. After three years of early cutting, no reduction in plant species diversity was observed. The second cut should be postponed to July to enable seed shed of plants. Grassland management decisions to control toxic C. autumnale must be made in close cooperation with nature conservation authorities to consider site characteristics and requirements of endangered species.

Key words

• biodiversity
• conservation
• control
• cutting date
• cutting programmes
• demography
• domestic animals
• dormancy
• ecology
• effects
• endangered species
• farmers
• feeds
• fodder
• grassland management
• grasslands
• growth rate
• habitats
• hay
• livestock
• marketing
• models
• natural grasslands
• nature conservation
• poisonous plants
• population dynamics
• population growth
• site class assessment
• species
• species diversity
• survival
• weeds