Investigating the ecological consequences of historic forest drainage through vegetation composition across Sweden

Abstract For hundreds of years, humans have drained water from landscapes, especially peatlands, to improve agriculture and forestry. Widespread, intensive ditching alters ecosystem processes and vegetation communities in ways that may last decades after ditch abandonment and that could be irreversible or require extensive intervention to reverse. However, the extent and nature of these legacy effects at broad spatial scales remain poorly understood. To investigate legacy effects of ditching on vegetation and soils, we used the entire country of Sweden as a case study because it has >150 years of drainage history and an extensive drainage network spanning >1 million kilometers. By combining a recently developed, high‐resolution ditch map with data from 4553 national forest and soil inventory sample plots, we compared ditched and non‐ditched plots across peat and mineral soils. We used ordinations and Bayesian regression models and found widespread evidence of legacy effects. Vascular plant diversity was higher on ditched plots, particularly on peat soils (31% higher than non‐ditched), and vegetation communities on drained peat approximated, yet remained distinct from, those on mineral soil and non‐ditched peat. Proximity to ditches, especially on peat, correlated with reduced variation in soil moisture and nutrients, more shade‐adapted species, and increased tree growth. Differences in community composition were associated with soil pH and C:N ratio, and especially ecological indicator values for nitrogen, pH, and moisture, suggesting that drainage has reshaped underlying environmental gradients. Our study therefore indicates that widespread drainage has left a legacy of transformed ecosystems. Given the scale and the magnitude of change, restoration to a previous ecological state may not be possible nor desirable in many areas. To help prioritize restoration efforts, species‐based indicators, especially those for nitrogen, pH, and moisture, may prove useful for a quick assessment of what sites have the most severe or irreversible alterations. Our results are critical for implementing large‐scale ecosystem restoration goals and making strategic decisions about which landscapes to restore or to manage as they are because they have breached ecological thresholds or developed desirable ecosystem services.