Mary F. Moffett

A bioassessment approach for mid-continent great rivers: the Upper Mississippi, Missouri, and Ohio (USA)

The objectives of the Environmental Monitoring and Assessment Program for Great River Ecosystems (EMAP-GRE) are to (1) develop and demonstrate, in collaboration with states, an assessment program yielding spatially unbiased estimates of the condition of mid-continent great rivers; (2) evaluate environmental indicators for assessing great rivers; and (3) assess the current condition of selected great river resources. The purpose of this paper is to describe EMAP-GRE using examples based on data collected in 2004–006 with emphasis on an approach to determining reference conditions. EMAP-GRE includes the Upper Mississippi River, the Missouri River, and the Ohio River. Indicators include biotic assemblages (fish, macroinvertebrates, plankton, algae), water chemistry, and aquatic and riparian physical habitat. Reference strata (river reaches for which a single reference expectation is appropriate) were determined by ordination of the fish assemblage and examination of spatial variation in environmental variables. Least disturbed condition of fish assemblages for reference strata was determined by empirical modeling in which we related fish assemblage metrics to a multimetric stressor gradient. We inferred least disturbed condition from the y-intercept, the predicted condition when stress was least. Thresholds for dividing the resource into management-relevant condition classes for biotic indicators were derived using predicted least disturbed condition to set the upper bound on the least disturbed condition class. Also discussed are the outputs of EMAP-GRE, including the assessment document, multimetric indices of condition, and unbiased data supporting state and tribal Clean Water Act reporting, adaptive management, and river restoration.

Benthic Versus Planktonic Foundations of Three Lake Superior Coastal Wetland Food Webs

Abstract The structure of aquatic food webs can provide information on system function, trophic dynamics and, potentially, responses to anthropogenic stressors. Stable isotope analyses in a Lake Superior coastal wetland (Allouez Bay, WI, USA) revealed that the food web was based upon carbon fixed by phytoplankton. We report analyses of two Lake Superior coastal wetland food webs not based upon plankton, indicating that Allouez Bay is not the prototype. In Lost Creek and West Fish Creek wetlands (WI, USA), upper trophic levels appear to be supported by benthic, rather than planktonic, food web pathways. Also, zooplankton in Lost Creek and West Fish Creek wetlands are enriched in 13 C relative to periphyton, rather than 13 C-depleted as reported in most other systems. This appears to be a consequence of zooplankton feeding on detrital seston that is 13 C-enriched relative to phytoplankton. We suggest possible causes for the observed differences in food web structure among these wetlands, including hydrologic differences among the systems, and responses to nutrient enrichment among the dominant forms of primary producers, similar to those described in lakes.

An assessment of stressor extent and biological condition in the North American mid-continent great rivers (USA)

We assessed the North American mid-continent great rivers (Upper Mississippi, Missouri, and Ohio). We estimated the extent of each river in most- (MDC) or least-disturbed condition (LDC) based on multiple biological response indicators: fish and macroinvertebrate, trophic state based on chlorophyll a, macrophyte cover, and exposure of fish-eating wildlife to toxic contaminants in fish tissue (Hg, total chlordane, total DDT, PCBs). We estimated the extent of stressors on each river including nutrients, suspended solids, sediment toxicity, invasive species, and land use (agriculture and impervious surface). All three rivers had a greater percent of their river length in MDC than in LDC based on fish assemblages. The Upper Mississippi River had the greatest percent of river length with eutrophic status. The Ohio River had the greatest percent of river length with fish with tissue contaminant levels toxic to wildlife. Overall, condition indices based on fish assemblages were more sensitive to stress than macroinvertebrate indices. Compared to the streams in its basin, more of the Upper Mississippi and Missouri Rivers were in MDC for nutrients than the Ohio River. Invasive species (Asian carp and Dreissenid mussels) were less widespread and less abundant on the Missouri River than on the other great rivers. The Ohio River had the most urbanized floodplains (greatest percent impervious surface). The Missouri River had the most floodplain agriculture. The effect of large urban areas on river condition was apparent for several indicators. Ecosystem condition based in fish assemblages, trophic state, and fish tissue contamination was related to land use on the floodplain and at the subcatchment scale. This is the first unbiased bioassessment of the mid-continent great rivers in the United States. The indicators, condition thresholds, results, and recommendations from this program are a starting point for improved future great river assessments.

Macroinvertebrate assemblage response to urbanization in three mid-continent USA great rivers

We sampled macroinvertebrates in the benthos and on the surface of woody snags in three mid-continent USA great rivers, the Upper Mississippi River, the Lower Missouri River, and the Ohio River, all of which flow through several large urban areas. We defined urban and non-urban zones of each river based on mean percent impervious surface extracted from recent land cover data. We estimated that 28-36 % of the rivers, by length, was urbanized. Based on multivariate ordination, the overall structure of the great river assemblages was not different between urban and non-urban zones for either the benthos or snags in any river. Most taxa (82 %) did not exhibit a significant positive or negative effect of urbanization in either river. Many of the taxa that were responsive to zone type were relatively rare (river- and habitat-specific relative abundance < 5 %). For responsive taxa, the effects of urbanization were generally weaker on the Upper Mississippi and Ohio Rivers than on the Lower Missouri River and weaker for snag-surface assemblages than for the littoral benthos. Among the taxa that were more abundant in urban sections were several insect taxa generally considered intolerant of pollution. We speculate that urbanization, possibly due to nutrient enrichment and resulting increased food resources (algae, seston), may improve conditions for these taxa. There appeared to be a threshold in mean percent impervious surface for a great river subcatchment near 5 % above which urban-intolerant taxa were always rare and urban tolerant taxa were often abundant.

Multimetric Fish Indices for Midcontinent (USA) Great Rivers

Abstract We developed fish assemblage–based multimetric indices as indicators of ecological conditions for the lower Missouri, impounded upper Mississippi, unimpounded upper Mississippi, and Ohio rivers. Using data from 475 daytime electrofishing samples, we calculated 81 candidate fish metrics. We screened these metrics for their range and their responsiveness to a multimetric stressor gradient that included water chemistry, human disturbance, and landscape-scale indicators of human disturbance and stress. Each reach-specific great-river fish index (GRFIn) included 8–10 nonredundant metrics scored on a continuous scale from 0 to 10 (10 = good). No metrics were common to all four GRFIns; one metric (the number of minnow species) was common to three GRFIns. We determined the least disturbed conditions from the y-intercepts of the quantile regressions between the GRFIn scores and the corresponding stressor gradients. Ecological condition as indicated by GRFIn score varied among reaches; 44 ± 8% (95% confide...

Spatial distributions of biophysical conditions on the Ohio River

We examined longitudinal and patchy distributions of water quality, fish assemblage, and land cover metrics from 59 sites on the Ohio River to determine if they were responses to geophysical, biological, and/or anthropogenic constraints. Conceptual models of river ecology hold that longitudinal trends are associated with river size and spatial heterogeneity or patchiness results from interactions between biology and hydromorphology, and in-channel and landscape process. Spatial autocorrelation was used to estimate patch length or the river distance between sites at which conditions are no longer spatially dependent. Significant linear (p 400 km). The exceptionally long patch length of silica suggests its distribution was constrained more by advection than biological assimilation like other nutrients. Turbidity, total suspended sediments, and Secchi depth had similar patch lengths. Their distributions were not related to the positions of dams or tributaries. The relatively small patch lengths for the proportion of fish with deformities or of native species, or of nonindigenous species contrasted with the large patch length for the proportion of detritivores. Patch length of riparian development was less than agriculture and forest suggesting that the distribution of development was relative more constrained by valley walls and vulnerability to flooding. Analyses of the spatial distributions offish, habitat, and water chemistry in large rivers can inform restoration and conservation decisions that maintain or increase biological diversity.