Use of Airborne Digital Imagery to Examine Floodplain Complexity at Varying Discharges

Floodplains are composed of aquatic and terrestrial habitats that are frequently reshaped by hydrologic processes operating at various spatial and temporal scales. I hypothesized that floodplain habitat complexity is maximized at intermediate discharges because small changes in flow result in substantial aquatic habitat changes and extreme discharges are associated with a decreased habitat heterogeneity. Between April and September 2014, I collected ultra-high resolution digital multispectral imagery of the Clark Fork River, Montana taken on 6 dates between early spring and fall. Following image mosaicking into a single image, unsupervised classification of the spectral reflectance was used to identify and quantify different aquatic habitats observed in the main channel, parafluvial, and orthofluvial zones of the floodplain. Through the course of the seasonal flood pulse, I observed significant changes in the spatial abundance of many habitat cover types (riffles, runs, shallow shorelines, overbank flow), but not others (backwaters, springbrooks, pools, ponds), suggesting that discharge is only a partial driver of the abundance of aquatic habitats. Riffles and runs and the most common transitions from one habitat to another dominated the main channel over the hydrograph changes that occurred between these habitats. The dominance of these habitats among main channel habitats was reflected in the low alpha diversity of the main channel, which was least diverse during peak flow conditions. Additionally, the main channel generally had low beta diversity, indicating that plots were usually very similar in habitat composition. The parafluvial zone was dominated by cobbles at low flows, transitioned to isolated parafluvial flood channels at moderate discharges, and dominated the expanded main channel during peak flow conditions. The parafluvial was the most diverse zone with peaks in alpha diversity occurring at intermediate flows on both the rising and falling limbs of the hydrograph. Furthermore, the beta diversity of the parafluvial zone tended to be high, suggesting that parafluvial plots were often dissimilar from each other. The orthofluvial zone was dominated by herbaceous habitat; however, I did observe aquatic habitats in the orthofluvial as well. Orthofluvial springbrooks transitioned to flood channels during high discharge as their upstream end connected during elevated discharges and then returned to springbrooks after the flood. The orthofluvial zone had an intermediate level of alpha diversity with the largest habitat diversity observed during peak flow. The beta diversity values of this zone indicated that most plots had some habitat cover type in common. I concluded that there is a relationship between discharge and floodplain habitat complexity, however it is influenced by an interaction between location on the riverscape and discharge.