The 14 taxa included seven dry-grassland species and seven wetland species, and the historical material came from the Herbarium of the Faculty of Biology at the University of Warsaw. Researchers sought to relocate each historical collection site, searching the original location when descriptions were precise and expanding to a radius of no more than 10 kilometres when necessary. Each historical–contemporary pair was analysed for the nitrogen isotope composition of leaves, carbon isotope composition and total nitrogen content.

Contemporary samples consisted of two to five leaf fragments, usually from several individuals at a locality, and were collected from April to September 2024. Calibration precision was below 0.1‰ for δ13C and below 0.2‰ for δ15N. Historical carbon-isotope values were adjusted for the progressive depletion of atmospheric carbon-13 caused by fossil-fuel emissions, known as the Suess effect.

The initial comparison showed a persistent separation between the habitats. In contemporary samples, mean leaf δ15N was −1.91‰ in dry-grassland species and 5.73‰ in wetland species, a difference of 3.83‰. The historical means were 1.24‰ and 4.45‰, respectively, producing a smaller but still significant difference of 3.22‰.

Wetland foliage also contained more total nitrogen in both periods: contemporary means were 3.78% for wetlands and 2.72% for dry grasslands, compared with 3.47% and 2.92% in the historical samples. These contrasts establish that the habitats differed in nitrogen-isotope composition and tissue nitrogen content before as well as after the period of widespread fertiliser use. The most striking temporal result was the reversal in the direction of the δ15N change.

Contemporary dry-grassland samples had a mean δ15N difference of −3.14‰ relative to their historical counterparts, indicating significantly lower values in modern leaves. All seven dry-grassland taxa showed negative mean changes, although the difference for Carex flacca was not significant. Wetland species moved in the opposite direction, with a pooled increase of 1.28‰, but that increase was not statistically significant for any individual wetland species.

The wetland group also showed greater within-group variation in δ15N change, which limited the precision of species-level comparisons. Multiple linear regression identified habitat type and the surrounding agricultural landscape as significant factors in the historical–contemporary difference in δ15N. The difference was greater in wetlands than in dry grasslands and increased with the proportion of farmland within a 10-kilometre radius of the collection site.

Nitrogen oxide deposition was also positively related to the δ15N difference, although at a lower level of significance. For changes in total nitrogen, habitat type was the only significant predictor in the model. The observational design links these patterns to land use and atmospheric inputs, but it does not independently establish which nitrogen source caused each change.

Total leaf nitrogen did not increase consistently across all habitats. The researchers found no significant overall historical–contemporary difference when dry-grassland species were pooled, although individual taxa differed. Carex praecox had significantly less total nitrogen in contemporary samples, while Silene otites had significantly more.

Among wetland plants, the pooled mean rose by 0.31 percentage points, a significant increase. Cardamine amara, Elodea canadensis and Stuckenia pectinata each showed significant increases, and their relative rises exceeded 20%. The result supports an increase in leaf nitrogen among the sampled wetland taxa, rather than a general century-long rise in nitrogen content across all of the plants examined.

Carbon isotopes separated the habitats even more sharply. After historical samples were corrected for the Suess effect, contemporary leaves had significantly lower δ13C in both environments, but the mean changes were −0.51‰ for dry grasslands and −3.85‰ for wetlands. Five wetland taxa—Alisma species, Cardamine amara, Elodea canadensis, Sagittaria sagittifolia and Stuckenia pectinata—showed significant declines.

Only one dry-grassland taxon, Carex praecox, had a significant carbon-isotope change. The largest differences occurred in the submerged plants E. canadensis and S. pectinata, with mean changes of −8.38‰ and −7.55‰, respectively. The authors considered several possible mechanisms for the divergent patterns.

Dry grasslands may have been affected primarily by changes in atmospheric composition, including deposition of combustion-derived nitrogen that is generally depleted in nitrogen-15. The abandonment of grazing and the resulting disappearance of livestock dung from many Polish grasslands may also have removed a comparatively nitrogen-15-enriched source. Wetlands can receive nitrogen and carbon through sewage, manure, fertiliser runoff, peat disturbance and other organic inputs transported by water.

The simultaneous movement in wetland δ15N and δ13C was consistent with increased inputs of organic matter, although the study did not directly reconstruct the sources responsible for each paired change. Water conditions provide a further context for the contrast. Dry soils limit microbial processes that can enrich nitrogen-15, potentially allowing atmospheric deposition to influence plant nitrogen more directly.

Wetland environments support more extensive processing of organic matter, and their carbon-isotope values can also be influenced by dissolved inorganic carbon produced during microbial decomposition. The study’s additional carbonate check found no measurable effect for S. pectinata and only a minor possible effect for E. canadensis that did not change the overall interpretation. Nevertheless, changes in drainage, drought, water chemistry, ecosystem management and surrounding land use may have affected the plants alongside direct pollution inputs.

Several limitations constrain how broadly the findings can be applied. The study covered 14 selected C3 plant taxa in one country, excluded species that fix nitrogen through symbiotic bacteria, and used relatively few observations for each species. Contemporary plants were not always recoverable at the exact historical locations because wetlands had disappeared or their trophic state and surroundings had changed.

Paired comparisons were strongest for dry-grassland δ15N, while no individual wetland species showed a significant nitrogen-isotope increase. The results therefore document contrasting measured responses in selected Polish habitats rather than prove a single universal pathway of industrial pollution. They show that wetland plants carried the more pronounced carbon-isotope and total-nitrogen changes, whereas dry-grassland plants mainly recorded a decline in δ15N.

Subject of Research: Comparative stable-isotope analysis of historical and contemporary wetland and dry-grassland plants in Poland

Article Title: Contrasting isotopic responses of dryland and wetland plants to a century of global anthropogenic changes in nutrient cycling

Article References: Dembicz, I., Chojnowska, N., Chibowski, P., & Kozub, Ł. (2026). Contrasting isotopic responses of dryland and wetland plants to a century of global anthropogenic changes in nutrient cycling. Biogeosciences, 23(20), 7375-7396. https://doi.org/10.5194/bg-23-7375-2026

Image Credits: AI Generated

DOI: 10.5194/bg-23-7375-2026

Keywords: plant isotopes, wetland plants, dry grasslands, nutrient cycling, nitrogen isotopes, carbon isotopes, herbarium specimens, eutrophication, land use, Poland, Contrasting, isotopic