Field study finds variable vegetation boundaries at Hungarian common milkweed invasion fronts

A 2023 field study of six common milkweed invasion fronts in Hungarian sand grasslands found that the visible plant boundary did not consistently coincide with statistically significant changes in species composition. Depending on the transect and the dissimilarity measure used, compositional boundaries occurred before, beyond or close to the last observed common milkweed shoot, with separations of up to 3.2 metres.

The study examined Asclepias syriaca, a clonal, rhizomatous perennial native to North America that is spreading in Central Europe. Investigators worked in the protected Fülöpháza Sand Dunes within Kiskunság National Park, in the Pannonic sand steppe of central Hungary. The landscape consists of calcareous sand dunes with open perennial grasslands, and all six sampled stands were located within semi-natural grassland categories, including Natura 2000 habitat.

Field sampling took place in May and June 2023 at six spatially separated invaded stands selected to represent differences in common milkweed shoot density and invasion pattern. Sites affected by roads, wheel tracks or intense trampling were excluded, and transects avoided major topographic variation. The transects were deliberately positioned rather than randomly placed so that each crossed a clearly identifiable invasion front and included invaded, non-invaded and transitional vegetation. Because the stands shared a regional invasion context, the authors treated the transects as spatial replicates within that context rather than fully independent invasion events.

Each transect was 12 metres long and 1 metre wide, with the central 6 metres containing the last detected common milkweed shoots. The sampling design divided every transect into 1,200 contiguous quadrats, each measuring 10 cm by 10 cm. Investigators recorded the presence of vascular plants, mosses and terricolous lichens, then aggregated the records across the transect width into 10-centimetre segments. Moving split-window analysis compared species composition in adjacent half-windows whose widths ranged from 10 to 100 centimetres.

Two dissimilarity measures characterized the boundaries. Squared Euclidean Distance was sensitive to differences in species frequencies, while the Differential Renkonen Index captured relative changes in species dominance. Restricted random-shift permutations with 1,000 Monte Carlo iterations generated null models used to standardize the profiles. Peaks above a threshold of Z = 1.65, corresponding to a one-tailed p = 0.05, were classified as statistically significant boundaries. Boundary position, peak height and peak width were used to distinguish sharp transitions from more diffuse zones, although boundary curvature could not be estimated along the linear transects.

Common milkweed was consistently well established, with 42–64 shoots recorded per transect, while total species richness ranged from 21 to 38. Its abundance was not associated with a clear reduction in richness within the sampled transects. Transects with relatively high milkweed counts also contained numerous other abundant taxa, including Secale sylvestre, Stipa borysthenica and the moss Tortula muralis. The observed pattern was therefore chiefly a change in dominance and community configuration rather than a demonstrated loss of species.

Squared Euclidean Distance detected significant boundaries in all six transects, but their positions and structures varied. In Transect 3, significant peaks occurred from 3.9 to 4.2 metres, whereas the last milkweed shoot was recorded at 7.4 metres, creating a 3.2-metre separation. The main dissimilarity zone in Transect 5, at 3.6–3.8 metres, preceded its 6.6-metre shoot limit by 2.8 metres. In Transect 2, the significant maximum lay 0.7 metres beyond the invasion front, while Transect 4 showed close alignment between the last shoot and several compositional peaks.

The Differential Renkonen Index produced a different and sometimes additional pattern. It detected no significant maximum in Transects 1 and 5. Transect 4 contained significant zones at 2.1–2.4 metres and 3.6–3.8 metres, both before its 6.5-metre invasion front, while Transect 3 had a maximum at 9.0 metres, 1.6 metres beyond the last milkweed shoot. In Transect 6, a continuous significant zone from 6.1 to 6.8 metres contained the visible front at 6.3 metres. These contrasting results show that the detected boundary depended on whether the analysis emphasized species-frequency differences or changes in relative dominance.

Overall, the findings support spatially structured transition zones associated with common milkweed, but not a uniform sequence in which ecological change always precedes visible colonization. The authors considered possible explanations including invasion stage, community resistance, legacy effects, microclimate, competition and below-ground interactions, none of which was directly tested. The analysis was limited by six deliberately positioned transects in one habitat type and region, a single field season and the absence of site-specific invasion histories, so it cannot establish invasion rates or direction over time. Detailed soil-moisture, microtopography, disturbance-history and below-ground measurements were also unavailable, preventing biotic effects from being separated from fine-scale environmental variation. The authors considered the labor-intensive method more suitable for targeted ecological assessment and management planning than for rapid, large-scale early detection.

Subject of Research: Fine-scale vegetation boundary detection at common milkweed invasion fronts in Hungarian sand grasslands

Article Title: Hidden fronts of invasion: Fine-scale boundaries reshape native plant communities

Article References: Bakacsy, L. (2026). Hidden fronts of invasion: Fine-scale boundaries reshape native plant communities. NeoBiota, 107, 287-307. https://doi.org/10.3897/neobiota.107.185747

Image Credits: AI Generated

DOI: 10.3897/neobiota.107.185747

Keywords: common milkweed, plant invasion, vegetation boundaries, novel ecotones, species composition, Hungary, sand grasslands, dissimilarity analysis, field study, Hidden, fronts, invasion