The double-edged sword of agricultural tile drainage effects on stream flashiness: A meta-analysis

• Meta-analysis of tile drainage impacts on stream flashiness in 464 watersheds worldwide. • In flat topography watersheds, stream flashiness increased with more tiling. • In sloped topography watersheds, stream flashiness increased with less tiling. • Watersheds with low infiltration soils had + 24 % stream flashiness when less tiled. • Stream flashiness in watersheds with high infiltration soils unaffected by tile drainage. Tile drainage is performed to increase agricultural productivity and can significantly alter downstream hydrology. A comprehensive understanding of tile drainage impacts on stream flashiness can inform tile drainage design and management at the watershed scale to balance agronomic benefits and potential impacts on watershed hydrology. We evaluated by meta -analysis the impact of tile drainage on downstream flashiness (Richards-Baker Index, RBI) for 464 rain-fed watersheds worldwide encompassing highly variable tile drainage extent and watershed physical characteristics, identified by systematic literature review and USGS datasets. Watersheds were largely concentrated in North America (89.9 %), particularly the US Upper Midwest (77.4 %), and northern Europe (9.7 %). Stream flashiness was similar regardless of tile drainage extent. However, stream flashiness was higher in relatively flat (2 %, flashiness was greater for low tile drainage (RBI 0.49) than medium (RBI 0.25) and high (RBI 0.35) tile drainage. Watersheds with low infiltration soils exhibited 24 % greater flashiness for low (RBI 0.51) tile drainage extent than for high (RBI 0.41) tile drainage extent. In watersheds with soils of high infiltration rate, stream flashiness was similar across the low (RBI 0.28), medium (RBI 0.30), and high (RBI 0.29) tile drainage extents. By accounting for watershed characteristics of slope and soil infiltration, this work points to complex drivers of tile drainage outcomes for stream flashiness that can lead to opposite impacts of tile drainage on stream flashiness and thus preclude blanket conclusions of tiling effects on watershed hydrology. Additionally, we identify salient data gaps in tile system design and management (e.g., tile diameter, installation depth, and spacing) that must be resolved to enable more accurate assessments of context-specific tile drainage impacts on stream flashiness.