Data from: Improving estimates of environmental change using multilevel regression models of Ellenberg indicator values

Ellenberg indicator values (EIVs) are a widely used metric in plant
ecology comprising a semi-quantitative description of species‘ ecological
requirements. Typically, point estimates of mean EIV scores are compared
to infer differences in the environmental conditions structuring plant
communities – particularly in resurvey studies with no historical
environmental data available. However, the use of point estimates as a
basis for inference does not take into account variance among species EIVs
within sampled plots, and gives equal weighting to means calculated from
sites with differing numbers of species. We present a set of multilevel
models – fitted with and without group-level predictors – to improve
precision and accuracy of site mean EIV scores, and to provide more
reliable inference on changing environmental conditions over spatial and
temporal gradients in re-visitation studies. We compare multilevel model
performance to GLMM’s fitted to point estimates of site mean EIVs. We also
test the reliability of this method to improve inferences with incomplete
species lists in some or all sample sites. Hierarchical modelling led to
more accurate and precise estimates of site-level differences in mean EIV
scores between time-periods, particularly for datasets with incomplete
records of species occurrence. They also revealed directional
environmental change within ecological habitat types, which estimates from
GLMM’s were inadequate to detect. Multilevel models also highlighted a
prominent role of hydrological differences as a driver of community change
in our case study, which traditional use of EIVs failed to reveal. We have
demonstrated that multilevel modelling of EIVs allows for a nuanced
estimation of environmental change underlying ecological communities from
plant assemblage data, leading to a better understanding of temporal
dynamics of ecosystems. Further, the ability of these methods to perform
well with missing data should increase the total set of historical data
which can be used to this end.