Search

ABSTRACT Overbank flooding is thought to be a critical process controlling nitrogen retention and cycling. Yet, studies aimed at quantifying these effects, specifically nitrification, are relatively few. In this study, we investigated the effects of season and flood frequency on soil nitrification rates in forested floodplains of Upper Mississippi River, Pool 8. Samples were collected from three plots within each site in April, August and November 2006. Plots were equally divided among three flood frequency categories as follows: rare, moderate and frequent based on elevation and flood probability model. We found a significant difference in nitrification rates among flood frequency categories as follows: rare > moderate > frequent ( F = 4.49, p < 0.01) and over season: spring > summer > autumn ( F = 8.88, p < 0.01). Regression for all samples showed that elevation, NH4-N, bulk density and soil temperature explained a moderate amount of variation in nitrification rates ( R2 = 0.29, p < 0.01). Models for moderately flooded, spring, summer and autumn samples improved when analysed individually. The absence of a correlation between nitrification rates and hydrology limits our ability to predict rates based on hydrology alone. The model based on elevation and season allows us to estimate nitrification rates with moderate confidence ( R2 = 0.27, p < 0.01). A rough calculation of forest floodplain nitrification rates suggests that 473 mt of NO3-N are produced annually, about 0.5% of Pool 8 total annual NO3-N budget. Copyright © 2013 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

ABSTRACT We applied an Index of Biotic Integrity (IBI) used on Wisconsin/Minnesota waters of the upper Mississippi River (UMR) to compare data from two systemic sampling programmes. Ability to use data from multiple sampling programmes could extend spatial and temporal coverage of river assessment and monitoring efforts. We normalized for effort and tested fish community data collected by the Environmental Monitoring and Assessment Program-Great Rivers Ecosystems (EMAP-GRE) 2004-2006 and the Long Term Resource Monitoring Program (LTRMP) 1993-2006. Each programme used daytime electrofishing along main channel borders but with some methodological and design differences. EMAP-GRE, designed for baseline and, eventually, compliance monitoring, used a probabilistic, continuous design. LTRMP, designed primarily for baseline and trend monitoring, used a stratified random design in five discrete study reaches. Analysis of similarity indicated no significant difference between EMAP-GRE and LTRMP IBI scores ( n = 238; Global R = −0.052; significance level = 0.972). Both datasets distinguished clear differences only between 'Fair' and 'Poor' condition categories, potentially supporting a 'pass-fail' assessment strategy. Thirteen years of LTRMP data demonstrated stable IBI scores through time in four of five reaches sampled. LTRMP and EMAP-GRE IBI scores correlated along the UMR's upstream to downstream gradient (df [3, 25]; F = 1.61; p = 0.22). A decline in IBI scores from upstream to downstream was consistent with UMR fish community studies and a previous, empirically modelled human disturbance gradient. Comparability between EMAP-GRE (best upstream to downstream coverage) and LTRMP data (best coverage over time and across the floodplain) supports a next step of developing and testing a systemic, multi-metric fish index on the UMR that both approaches could inform. Copyright © 2011 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]