The research, published in the journal Ambio, surveyed 32 stream and ditch reaches within the Krycklan Catchment Study, a long-term research landscape about 60 kilometres northwest of Umeå in subarctic northern Sweden. All sites were headwaters with catchment areas below a 60-hectare threshold, stood in mature forest with trees taller than ten metres, and were dominated by Norway spruce and Scots pine.

Fieldwork took place in 2023. At each reach, the researchers surveyed a ten-metre section in October, taking three one-minute kick-net samples with 250-micrometre mesh across the available habitat types and combining them into a single composite sample per site. Macroinvertebrates were sorted under Swedish national assessment protocols, identified to the lowest taxonomic level reliably achievable, and assigned to functional feeding groups using the freshwaterecology.info database. Alongside the biological sampling, the team measured substrate composition on the Wentworth scale, instream vegetation, deadwood, water depth, canopy cover, slope, water velocity, pH, turbidity and total dissolved solids, with depth measured on three occasions across July, August and October.

The headline biological result was nuanced. Taxa richness did differ among waterway types: natural streams on till soils supported about 1.5 times more taxa than peat ditches and 1.4 times more than till ditches, a statistically significant difference. EPT richness, the number of mayfly, stonefly and caddisfly taxa widely used as indicators of freshwater condition, was also higher in till streams and modified waterways than in till ditches. Yet two broader measures told a different story. Overall macroinvertebrate abundance and Shannon diversity showed no significant differences among the four waterway types, and the number of taxa observed per site ranged from six to eighteen across the whole dataset.

Community composition added further texture. Non-metric multidimensional scaling of presence-absence and abundance data revealed groupings by waterway type, with partial separation between peat-based and till-based systems, and analysis of similarities confirmed significant differences among all four categories. But when the analysis was restricted to the till-based waterways only, the pattern weakened: communities overlapped, the separation was not statistically significant, and only catchment area correlated with the ordination. The authors interpret this as possible ecological convergence, with ditch and modified-channel communities on mineral soils gradually approaching those of natural streams as physical conditions stabilise over time, though the moderate ordination stress values mean the patterns should be read with caution.

The clearest environmental divide ran along soil lines rather than channel origin. Peat ditches carried nearly double the organic substrate cover of till ditches and more than four times that of till modified waterways and natural streams, while holding 13.5 times less medium-sized substrate than modified waterways and 45 times less coarse substrate than natural streams. They also had the lowest pH, consistently below 5, a recognised threshold below which ion regulation becomes physiologically difficult for many macroinvertebrates and aluminium toxicity rises. Peat ditches additionally showed the slowest water velocities, roughly seven times lower than till streams, the most instream vegetation, and higher turbidity and dissolved solids. All till-based waterways, by contrast, remained chemically similar to one another and above the acidic threshold.

When the team modelled which factors best predicted taxa richness across the full gradient, two variables survived model selection: pH and catchment area. Both had positive effects, with taxa richness increasing significantly with each, and the final model explained roughly 49 percent of the variation in richness. Catchment area was also correlated with habitat structure, since larger catchments had coarser substrate, faster flow and greater depth, while smaller headwaters supported more instream vegetation. Functional feeding groups varied only modestly: gatherer-collectors and shredders showed no significant differences among types, passive filter feeders, dominated by Simuliidae, were about five times more abundant in till modified waterways than in till ditches, and predators were proportionally more common in till ditches, likely reflecting taxa adapted to low-oxygen, slow-water conditions.

These findings carry weight because Swedish forestry law treats the different waterway categories very differently. Forests managed for timber cover 56 percent of Sweden’s land area, and clear-cutting is the prevailing harvest method. Riparian buffers are typically left along natural and modified streams, averaging 7.9 and 4.6 metres on either side respectively, but ditches are exempt from such requirements and are commonly dredged and cleaned to maintain drainage capacity. At the same time, ditch infilling to rewet peatlands is increasingly promoted for climate and ecological reasons, creating trade-offs with timber production. If ditches on mineral soil genuinely support stream-like communities, the study’s authors argue, blanket management rules based on perceived naturalness may misjudge where biodiversity actually resides.

The comparison with earlier drainage research reinforces the point. Previous work in boreal headwaters found that sites with 0 to 35 percent of catchment drained averaged four to six EPT taxa, while areas with 35 to 50 percent drained averaged about two. The streams and modified waterways in the Krycklan study resembled the less-drained end of that spectrum, while the ditches resembled the more heavily drained end. The authors suggest that smaller, more isolated peat ditches with strong chemical constraints may be candidates for blocking and peatland restoration, whereas larger, better-connected till ditches and modified waterways could be prioritised for stream restoration instead.

The study has limits the authors acknowledge. It covers a single boreal catchment, sampling occurred in one autumn season, and some identifications could only be resolved to genus, family or order. The convergence interpretation remains a hypothesis that would need chronosequences of waterways of differing ages since construction or last cleaning to confirm. Still, the baseline matters: artificial waterways make up most of Sweden’s headwater network, and the finding that modified channels and many ditches can host biodiversity comparable to natural streams suggests they deserve consideration in conservation and restoration planning, and possibly similar protections, rather than being written off as ecological dead ends.

Subject of Research: Benthic macroinvertebrate biodiversity in natural, modified and artificial waterways of a boreal Swedish forest catchment

Article Title: Small waterways in boreal forests: How artificial waterways influence benthic macroinvertebrate communities

Article References: Zivec, P., Sponseller, R., Kuglerová, L., Laudon, H., Trail, A., & Hasselquist, E. M. (2026). Small waterways in boreal forests: How artificial waterways influence benthic macroinvertebrate communities. Ambio. https://doi.org/10.1007/s13280-026-02488-7

Image Credits: AI Generated

DOI: 10.1007/s13280-026-02488-7

Keywords: boreal forests, ditches, macroinvertebrates, stream ecology, biodiversity, Sweden, Krycklan Catchment Study, peatlands, pH, catchment area, restoration, forestry