Umeå universitets logga

River systems form dendritic ecological networks that influence the spatial structure of riverine communities. Few empirical studies have evaluated how regional, dispersal-related processes and local habitat factors interact to govern network patterns of species composition. We explore such interactions in a boreal watershed and show that riparian plant species richness increases strongly with drainage size, i.e., with downstream position in the network. Assemblage composition was nested, with new species successively added downstream. These spatial patterns in species composition were related to a combination of local and regional processes. Breadth in local habitat conditions increased downstream in the network, resulting in higher habitat heterogeneity and reduced niche overlap among species, which together with similar trends in disturbance, allows more species to coexist. Riparian edaphic conditions were also increasingly favorable to more species within the regional pool along larger streams, with greater nitrogen availability (manifested as lower C:N) and more rapid mineralization of C and N (as indicated by ratios of stable isotopes) observed with downstream position in the network. The number of species with capacity for water dispersal increased with stream size providing a mechanistic link between plant traits and the downstream accumulation of species as more propagules arrive from upstream sites. Similarity in species composition between sites was related to both geographical and environmental distance. Our results provide the first empirical evidence that position in the river network drives spatial patterns in riparian plant diversity and composition by the joint influence of local (disturbance, habitat conditions, and breadth) and regional (dispersal) forces.

Riparian vegetation research has traditionally focused on channel-related processes because riparian areas are situated on the edge of aquatic ecosystems and are therefore greatly affected by the flow regime of streams and rivers. However, due to their low topographic position in the landscape, riparian areas receive significant inputs of water and nutrients from uplands. These inputs may be important for riparian vegetation, but their role for riparian plant diversity is poorly known. We studied the relationship between the influx of groundwater (GW) from upland areas and riparian plant diversity and composition along a stream size gradient, ranging from small basins lacking permanent streams to a seventh-order river in northern Sweden. We selected riparian sites with and without GW discharge using a hydrological model describing GW flow accumulation to test the hypothesis that riparian sites with GW discharge harbor plant communities with higher species richness. We further investigated several environmental factors to detect habitat differences between sites differing in GW discharge conditions. Vascular plant species richness was between 15% and 20% higher, depending on the spatial scale sampled, at riparian sites with GW discharge in comparison to non-discharge sites, a pattern that was consistent across all stream sizes. The elevated species richness was best explained by higher soil pH and higher nitrogen availability (manifested as lower soil C/N ratio), conditions which were positively correlated with GW discharge. Base cations and possibly nitrogen transported by groundwater may therefore act as a terrestrial subsidy of riparian vegetation. The stable isotopes N-15 and C-13 were depleted in soils from GW discharge compared to non-discharge sites, suggesting that GW inputs might also affect nitrogen and carbon dynamics in riparian soils. Despite the fact that many flows of water and nutrients reaching streams are filtered through riparian zones, the importance of these flows for riparian vegetation has not been appreciated. Our results demonstrated strong relationships between GW discharge, plant species richness and environmental conditions across the entire stream size gradient, suggesting that both river hydrology and upland inputs should be considered to fully understand riparian vegetation dynamics.

The distribution of water across landscapes affects the diversity and composition of ecological communities, as demonstrated by studies on variation in vascular plant communities along river networks and in relation to groundwater. However, nonvascular plants have been neglected in this regard. Bryophytes are dominant components of boreal flora, performing many ecosystem functions and affecting ecosystem processes, but how their diversity and species composition vary across catchments is poorly known. We asked how terrestrial assemblages of mosses and liverworts respond to variation in (i) catchment size, going from upland-forest to riparian settings along increasingly large streams and (ii) groundwater discharge conditions. We compared the patterns found for liverworts and mosses to vascular plants in the same set of study plots. Species richness of vascular plants and mosses increased with catchment size, whereas liverworts peaked along streams of intermediate size. All three taxonomic groups responded to groundwater discharge in riparian zones by maintaining high species richness further from the stream channel. Groundwater discharge thus provided riparian-like habitat further away from the streams and also in upland-forest sites compared to the non-discharge counterparts. In addition, soil chemistry (C:N ratio, pH) and light availability were important predictors of vascular plant species richness. Mosses and liverworts responded to the availability of specific substrates (stones and topographic hollows), but were also affected by soil C: N. Overall, assemblages of mosses and vascular plants exhibited many similarities in how they responded to hydrological gradients, whereas the patterns of liverworts differed from the other two groups.

Riparian forests (RFs) along streams and rivers in forested landscapes provide many ecosystem functions that are important for the biodiversity and biogeochemistry of both terrestrial and aquatic ecosystems. In riverine landscapes, many of these ecological and biogeochemical functions have been found to be maximized in riparian areas with discharge of upland-originating groundwater (GW). This ecological significance, and the fact that riparian areas with GW discharge are important sources of many chemical elements in streams and rivers, makes these places important hotspots in the landscape. The natural functioning of RFs is however threatened by poorly designed management practices, with forestry being one of the most important examples in timber producing regions. Logging operations in riparian, but also in adjoining upland forests, threaten to alter many riparian functions. This effect is accelerated in GW discharge hotspots because of their sensitive soils and the high connectivity with uphill areas. We thus argue that forestry practices should give higher consideration to riparian GW discharge areas, and we demonstrate how improved riparian buffer zone management can be incorporated into every-day forestry planning. We offer a practical tool for more optimized site-specific riparian buffer design by using model-derived high resolution maps with detailed information about wetness and soil–water flow paths within RFs. We describe how such site-specific riparian buffer management differs from fixed-width buffers, which are generally applied in today’s forestry, and address some risks connected to fixed-width buffer management. We conclude that site-specific riparian management, allowing wider buffers at GW discharge areas and more narrow buffers on sites of lower ecological significance (i.e. riparian sites without GW flow paths), would benefit a variety of ecosystem services, mitigate negative effects caused by forestry and create more variable and heterogeneous riparian corridors. Finally, we show examples of how the new forestry planning can be applied.

We evaluated the ecological effects of stream restoration by comparing riparian vascular plant diversity and riparian habitat properties between channelized stream reaches and reaches restored with two different techniques in northern Sweden. Channelized streams were modified >50 years ago to facilitate timber floating which implied straightening and clearing channels from in-stream boulders and wood, which were deposited in riparian areas. Traditionally restored reaches underwent restoration efforts 8-10 years ago when material from the riparian zones was returned back to the channels. Some reaches were re-restored 3 years before our study by bringing large structural elements (boulders and downed trees) to the previously traditionally restored streams. These three restoration types (i.e., channelized, traditionally restored, and re-restored) represent a gradient in habitat complexity and we tested whether they also represent a gradient in plant diversity, and habitat quantity and quality. Riparian plant cover was higher at both restored types than channelized reaches. Responses of riparian plant diversity to restoration were scale dependent, with higher species richness at both restored types than channelized reaches at plot scale level (0.25 m²), whereas at the reach (>700 m² of riparian area) and transect scales, plant species richness did not differ among the three restoration types. Differences in plots scale richness were largest in the most dynamic and species-rich part of riparian zones, i.e., at 20 and 40 cm elevations above the low summer water levels. We found no systematic differences in plant cover and species richness between the two restored types. Species composition did not differ among the restoration types at the reach scale, but at the plot scale species composition segregated between channelized and restored reaches but with small compositional differences between the traditionally restored and re-restored types. Aspects of both riparian habitat quantity and quality for vascular plants increased after restoration. Riparian zones along re-restored reaches had the highest amount of soils available for plant establishment in comparison to traditionally restored and channelized streams. In contrast, soil quality was more favorable to plants along channelized reaches, with higher soil pH and lower C:N ratio at some riparian levels. The magnitude of floods did not differ between the three types of streams, but inundation duration significantly increased following restoration with the longest inundation events at re-restored reaches. Thus, given that inundation is positively associated with plant species richness in riparian zones, habitat quality may have improved following restoration. We conclude that restoration of channelized streams has improved the abundance and diversity of riparian vascular plants. Given that we observed several positive trajectories in biotic and habitat recovery (although statistically insignificant in some cases) and that the new re-restoration technique is predicted to enhance riparian habitat conditions more effectively than traditional methods, we suggest that with time, further recovery of riparian vegetation may occur.