On January 15, 2026, I gave a talk at the Division of BioInvasions, Global Change & Macroecology, presenting the current state of my second PhD chapter. The talk, titled “Reconstructing the Patterns of Alien Plant Species Habitat Niche Expansion,” addressed the question of whether alien species expand into new habitat types over time and what drives this expansion.

Motivation

The regional accumulation of alien species is well documented, but how this translates to local-scale habitat dynamics remains less explored. While thousands of alien plants are recorded across European territories, most are concentrated in a small number of disturbed habitats. The central question was whether species with longer residence times progressively colonize additional habitat types, and if so, which traits and pathways characterize this expansion.

A New Overrepresentation Framework

A key methodological contribution of the talk was the overrepresentation framework developed to handle the opportunistic sampling structure of the European Vegetation Archive (EVA). The database contains 1.9 million vegetation plots from 56 countries spanning nine decades, but habitats are sampled unevenly. Standard diversity indices and fidelity measures assume systematic sampling and break down when applied to such data.

The framework reorganizes the data into hexagonal grid cells, each covering 863 km², crossed with 10-year time windows. Within each cell, the species’ habitat distribution is compared against the local background distribution of all plots. A species is classified as overrepresented in a habitat when its occurrence exceeds what would be expected given local availability. This approach provides a clear null model, controls for local sampling differences, and produces a categorical output directly usable as a response variable.

Hurdle Model for Habitat Breadth

The analysis used a two-stage hurdle model. The first stage addressed whether a species can extend beyond a single habitat type at all, distinguishing specialists from potential generalists. The second stage modelled how many additional habitats a species occupies once it crosses that threshold.

Key Findings

The results showed a clear residence time effect. Species present for fewer than 20 years had a 17% probability of occupying more than one habitat, rising to 57% for species with over 200 years of residence. Expected habitat counts increased from 1.21 to 2.04 across these cohorts.

Trait effects revealed an interesting pattern. Trees, vines, and long-lived perennials were less likely to cross the one-habitat threshold. However, long-lived perennials that did expand occupied 43% more habitats than expected, suggesting that once these species overcome initial barriers, they spread broadly.

At the habitat level, man-made and ruderal habitats served as the primary invasion gateway, accounting for 31% of first-habitat occurrences. Expansion pathways typically led from these disturbed habitats into dry grasslands, then to mesic and wet grasslands. Alpine grasslands and bogs showed the highest resistance to alien establishment.

Discussion

The talk concluded with a discussion of caveats, including the cross-sectional design (cohort comparison rather than tracking individual species over time), the absence of abundance data, and the continental aggregation of transition pathways. The division provided constructive feedback, particularly on refining the transition analysis and strengthening the link between residence time and mechanistic processes.