Two different UK limestone grasslands were exposed to simulated climate change with the use of nonintrusive techniques to manipulate local climate over 5 years. Resistance to climate change, defined as the ability of a community to maintain its composition and biomass in response to environmental stress, could be explained by reference to the functional composition and successional status of the grasslands. The more fertile, early-successional grassland was much more responsive to climate change. Resistance could not be explained by the particular climates experienced by the two grasslands. Productive, disturbed landscapes created by modern human activity may prove more vulnerable to climate change than older, traditional landscapes.

Summary 1.Changes in the frequency of extreme events, such as droughts, may be one of the most significant impacts of climate change for ecosystems. Models predict more frequent summer droughts in much of England: this paper investigates the impact on different types of plants in an ex-arable grassland community. 2. A long-term experiment simulated increased and decreased summer precipitation. Substantial interannual variation allowed the effects of summer drought to be tested in combination with wet and dry weather in other seasons. This is important, as climate models predict increased winter precipitation. 3. Total cover abundance in early summer increased with increasing water supply in the previous summer; there was no effect of winter precipitation. Productivity is therefore likely to decrease with more frequent summer droughts, with no mitigating effect of wetter winters. 4. The percentage cover of perennial grasses declined during a natural drought in 1995-97; this was exacerbated by the experimental drought treatment and reduced by supplemented rainfall. Simultaneously, short-lived ruderal species increased; this was greatest in drought treatments and least with supplemented rainfall. 5. These trends were subsequently reversed during several years of unusually wet weather, with perennial grasses increasing and short-lived forbs decreasing. This occurred even in experimentally droughted plots, and we propose that it resulted from rapid coverage of gaps during wet autumns and winters. 6. Deep-rooted species generally proved to be more drought resistant, but there were exceptions. 7. We conclude that increased frequency of summer droughts could have serious implications for the establishment and successional development of ex-arable grasslands. Increased winter precipitation would moderate the impact on species composition, but not on productivity.

1. Novel manipulations of local climate were employed to investigate how warmer winters with either wetter or drier summers would affect the Auchenorrhyncha, a major component of the insect fauna of grasslands. Direct and indirect effects of climate manipulation were found. 2. Supplemented summer rainfall resulted in an increase in vegetation cover, leading to an increase in the abundance of the Auchenorrhyncha. 3. Summer drought, however, caused a decrease in vegetation cover, but this did not lead to a corresponding decrease in the abundance of the Auchenorrhyncha. 4. Egg hatch and the termination of nymphal hibernation occurred earlier in winter warmed plots; however, the rate of nymphal development was unaffected.

1. One of the climate change scenarios predicted for the UK is warmer winters and additional summer rainfall, which may favour growth and survival of fungal pathogens. We tested several hypotheses on the fate of persistent seeds in the soil and the role of fungal pathogens under this predicted climate change. 2. We buried seed bags containing fungicide‐treated and non‐fungicide‐treated seeds of four species with persistent seed banks (Convolvulus arvensis L., Lotus corniculatus L., Medicago lupulina L. and Rubus fruticosus L.) under control and simulated climate change (winter warming plus supplemented summer rain) conditions, and monitored seed survival over 1 to 2 years. 3. Fungicide treatment resulted in a significant increase in the percentage of intact seeds recovered for only two of the four species, M. lupulina and R. fruticosus. Seeds of M. lupulina that were treated with fungicide remained viable in the soil for longer than non‐treated seeds. Thus, the effect of fungal pathogens on seed persistence in the soil appears to be species specific. 4. There was no significant effect of the simulated climate (winter warming plus supplemented summer rain) on seed persistence in the soil, for any of the four species. Neither was a significant climate × fungicide treatment interaction found for any of the four species. Thus, it does not appear that the conditions provided in the simulated climate plots favoured the growth and survival of fungal pathogens affecting the soil seed banks of the four species studied here. 5. The use of fungicides in manipulative experiments and the importance of field experiments that simulate predicted climate change are discussed.