Declines in insect biomass and individual numbers have been demonstrated in many parts of the world, including central European dry grasslands. It is, howver, unclear if biomass data or individual numbers are superior as measures of change in species composition, and in trait composition of insect communities. We revisited a former study of ours demonstrating severe losses in abundance of planthoppers and leafhoppers in Central European grasslands since the 1960s. We performed a series of univariate, multivariate and trait-based tests to investigate the relationship between biodiversity change and environment, and compared the results for number-weighted and biomass-weighted community data. While both individual numbers and biomass declined strongly over time, no changes in species diversity (Jost's D) were observed irrespective of the measure used for weighting. Surprisingly, metrics for trait diversity tended to increase over time in both the number-weighted and biomass-weighted data. For both measures, community composition changed strongly over time for species and trait data, and was also associated with landscape structure and mean annual precipitation. Of the tested traits, we only found associations between an early phenology and time, and a temporal increase of species over-wintering in a nymphal stage, but again there was no major difference between the measure used for weighting. Our study shows that declining assemblage size (regardless of the measure used) does not necessarily translate in concomitant changes in functional diversity or trait diversity. From a methodological perspective, no clear evidence emerged for either biomass or individual number as measure of assemblage size being more sensitive to environmental change. This implies that conversion of widely used data of individual numbers to biomass values is unlikely to bring much additional insights. With respect to the general scarcity of quantitative data on insect communities, the focus should be on using whatever data is available.