WOOD MOBILITY AS A CONTROL ON HYDRAULIC RESISTANCE IN FORESTED HEADWATER STREAMS

Instream wood, or large woody debris (LWD), contributes significantly to hydraulic flow resistance in certain geomorphic and hydrologic situations, while in other sites with similar wood loads it provides a relatively minor contribution. For example, we observe flow resistance that is easily accounted for by grain roughness alone in steep, boulder-bedded, Costa Rican study streams with large wood loads, while total resistance greatly exceeds grain resistance in nearby sandy-bedded, low-gradient, high-wood-load sites. And again in contrast to the steep, tropical Costa Rican streams, we observe hydraulic resistance much higher than that predicted by bed grain size alone in similarly steep, coarse-grained, wooded sites in Frasier Experimental Forest, Colorado, which has a snowmelt dominated flow regime.

We hypothesize that hydraulic resistance contributions from wood are controlled not by wood load alone, but also by wood mobility. Wood mobility is related to the relative size of the wood: wood diameter relative to stream depth, and wood length relative to stream width. However we would emphasize that stream depth and width should be considered at an elevated discharge, such as bankfull or the mean annual flood, rather than at base flow. Thus if slope, grain size, and wood size are equal, hydrologic regimes that have more frequent or greater-magnitude flood events will have more mobile wood than more stable regimes. Mobile wood will then be reoriented into streamlined configurations and preferentially deposited in low-velocity zones on the channel margins, reducing the contribution of wood to total resistance.

To test this hypothesis, we calculated the difference between observed flow resistance and predicted flow resistance using a non-dimensional hydraulic geometry equation that accounts for grain size and slope at study sites across a range of hydrologic regimes from snowmelt dominated to tropical. We then compared this discrepancy, which should be an approximate measure of wood resistance, to surrogates for wood mobility including mean annual discharge, slope, bankfull depth, stream power, and 90^th percentile wood length and diameter, as well as other local physical metrics. Simple and multiple regressions suggest that wood mobility is indeed a control on hydraulic resistance in certain circumstances.