Spatial patterns of microbial activity
All three studied rivers represent second-order tributaries of the Danube and are located within a relatively small territory. In spite of this, the rivers strongly and significantly differ in climate, geology, physiography, properties of the riverbed soils, vegetation and, consequently, microbial communities. They all spring at middle altitudes (750 to 1100 m a.s.l.) in forests, but they differ in the landscape character of their lower parts. A major part of the northernmost Kysuca river crosses inhabited areas; even though the immediate vicinity of the flow is frequently overgrown by trees and shrubs, broader surroundings are mostly dominated by agricultural land and artificial surfaces (CORINE land use types). The Turiec in the middle is surrounded mainly by grasslands and agricultural areas, while the valley of the southernmost Krupinica is the most forested of all streams with the least artificial surfaces in the vicinity. These differences in landscape characteristics are reflected in the soil properties of the riverbed. Kysuca and Turiec have neutral to slightly alkaline soils. This is not surprising for Turiec, as most of its tributaries come from carbonate rocks in the nearby mountains. However, the territory drained by Kysuca is the flysch zone of the Outer Carpathians, formed primarily of sandstones and shales. A high pH of the riverine soils is probably caused here by the wastewaters of the neighbouring settlements. Differences in nutrient levels and soil texture among the studied streams are also substantial.
The response of plant and microbial communities is not straightforward. On one hand, species richness and α-diversity levels of plants do not significantly differ among rivers. On the other hand, high β-diversity in all three catchments indicates a strong differentiation in the composition of plant communities. In the case of the microbial community, α-diversity was comparable to plants but β-diversity was much lower. In this context, three issues must be emphasized. First, as river corridors generally show a high floristic similarity to the surrounding landscape serving as a species pool for the colonization of the riverbanks (Renöfält et al. 2005a), a high variation of ecological conditions and landuse contributes to the floristic differentiation along the river. On the other hand, the colonization of riverbed soils by microbiota is mainly mediated by the waterflow (Tolkkinen et al. 2020). Second, the assessment of soil microbiota was limited to the organo-mineral A-horizon. However, microbial biomass and activity change along the vertical profile of soil: generally, they are the highest in the organic O-horizon and rapidly decrease towards mineral horizons (Gömöryová et al. 2013, 2020). The organic O-horizon, which is typical for forest soils and generally hosts the richest and most abundant microbial community, was missing at most plots and thus not considered. Finally, taxonomic diversity may be substantially higher than functional diversity reflecting the ability of microbial groups to metabolize different substrates (Bolger 2001; Zak et al. 1994). In any case, in spite of similar α-diversity levels, RDA demonstrated that microbial community composition differs between rivers, and is affected by both soil texture and chemistry (silt content, pH-KCl) and the local source of organic carbon, i.e. plant community.
In streams, geology and waterflow are expected to set a template defining microbial community composition, while the microbiota is responsible for the processes of organic matter transformation and transport at a given location (Ward et al. 2021). Hydrological connectivity is a primary factor driving biogeochemical activity in riparian corridors, as river water and its chemical and biological constituents are critical to the growth and metabolic functions of organisms (Dwivedi et al. 2021). The connectivity mediated by the waterflow is thus expected to result in spatial continuity of riverbed soil properties including the properties of soil microbiota. This expectation was fulfilled just partially. We did not expect continuity in vegetation characteristics, as plant community composition is strongly affected by neighbouring vegetation and less dependent on plant propagules brought by water. However, among soil properties, only soil and nitrogen content consistently correlated between neighbouring sites in Kysuca and Turiec, and only a few nutrient contents exhibited at least marginally significant Mantel correlations. As microbial activity was found to be primarily determined by the available resources (organic carbon, nitrogen, nutrients; cf. Supplementary materials; Table S2), it is not surprising that microbial activity parameters proved to be spatial continuous just in these two streams. In fact, we cannot properly explain why no spatial continuity was observed in Krupinica, which is the least human-affected out of all three rivers. However, in spite of missing similarity between neighbouring sites, overall trends of microbial activity were observed in the case of Krupinica and Kysuca: activity increased from the mouth towards the headwaters. A similar trend of microbial metabolic processes was observed in a Chinese river system around the Three Gorges reservoir (Yang et al. 2023). In our case, the trends were influenced by outlying high activity at the uppermost plots close to the river sources located in (semi)natural forests. A high cover of trees producing easily decomposable litter as a source of organic carbon and nitrogen (Acer spp., Fraxinus excelsior etc.) probably contributed to this spatial pattern.
Effects of invasive and alien plants
In contrast to overall plant species richness, the number of alien and invasive species significantly varied among the studied streams. Vegetation along the Turiec river was almost free of non-native elements, while the highest proportion of invasive plants was expectedly observed in Kysuca, which passes many settlements (up to 4 species on a plot), and surprisingly the highest mean number of alien species was observed in Krupinica crossing relatively natural landscapes, reaching up to 28 species on a plot near the river mouth, where the first settlements occur. This fact is probably connected with more favourable climatic conditions (higher air temperature and lower precipitation) and a less dynamic hydrological regime (lower flow rate) in the Krupinica river compared with the other two rivers (see Table 1). It is generally known that alien plant richness decreases along an altitudinal gradient (Seipel et al. 2016; Tomaseto et al. 2013). Moreover, plant invasiveness is known to change along the river flow (Renöfält et al. 2005b); this was confirmed also in our study, where the richness of both aliens and invasive plants increased near the river mouths.
Our study showed a significant effect of both alien and invasive species on the functional composition of soil microbiota, but no effects on microbial activity and diversity. The hitherto information about the effects of invasive plants on soil microorganisms is controversial and strongly depends on the studied ecosystem and the experimental setup. Custer and van Diepen (2020) concluded that plant invasion has just a minor impact on α-diversity of soil microorganisms. On the other hand, Ravit et al. (2003) observed decreased microbial richness and soil enzyme activity in invaded wetlands. Similarly, bamboo expansion in Chinese forests caused a decline of both α- and β- diversity of soil nematodes (Xiao et al. 2023). The problem of such studies is that they rely on manipulative or observational experiments with a single species or very few species, the comparability with our study is thus limited. Except Kysuca, the number of invasive species on experimental plots generally ranged from zero to one, which was not enough for a quantitative assessment of the effects of single species on the microbiota.