Grantee Research Project Results
Final Report: SEER: The Role of Natural Versus Anthropogenic Factors in Assessing Ecological Risk in Agricultural Watersheds
EPA Grant Number: R827784E02Title: SEER: The Role of Natural Versus Anthropogenic Factors in Assessing Ecological Risk in Agricultural Watersheds
Investigators: Watzin, Mary C. , Gotelli, Nicholas J. , Hoffmann, James
Institution: University of Vermont
EPA Project Officer: Chung, Serena
Project Period: November 1, 1999 through April 30, 2002
Project Amount: $268,633
RFA: EPSCoR (Experimental Program to Stimulate Competitive Research) (1999) RFA Text | Recipients Lists
Research Category: EPSCoR (The Experimental Program to Stimulate Competitive Research)
Objective:
The overall objective of this research project was to develop a conceptual framework for watershed systems that can be used to predict where agricultural disturbances pose a significant risk to the stream community and where restoration activities have the greatest likelihood of success. The specific objectives of this research project are to: (1) assess the influence of stream connectivity and adjacent land use on stream benthic macroinvertebrate and periphyton communities; (2) evaluate colonization processes in natural and agriculturally influenced stream reaches to determine which factors might be most important in assessing the potential for recovery in damaged stream systems; and (3) specifically evaluate the impacts of phosphorus and suspended sediments on stream periphyton and macroinvertebrates using a manipulative field experiment.
Summary/Accomplishments (Outputs/Outcomes):
Field sampling to address Objective 1 was completed in the summer of 2000. Samples were collected at 13 stream confluences throughout Vermont. Objectives 2 and 3 were addressed with manipulative experiments in the summer of 2001. To address Objective 2, specially configured sample collectors were deployed to sample drifting macroinvertebrates, macroinvertebrates moving through the hyporheic zone, emerging adults, and ovipositing adults in both agricultural and reference streams. To address Objective 3, a manipulative field experiment adding phosphorus, suspended sediments, and a combination of the two to experimental flumes deployed in a reference stream reach was conducted. Macroinvertebrates and periphyton were collected in the flumes over a 30-day experiment.
Sample sites in Year 1 of the project represented four combinations of adjacent agricultural and undisturbed land use. Each site included comparable stream segments extending from two second order streams through the tributary node to a third order stream reach. At each site, periphyton and macroinvertebrate samples were collected both downstream of the tributary juncture and upstream in each tributary reach. To analyze the data, we used a simple mixing model. The relative abundances of taxa in the upstream reaches were used to predict the relative contribution of each of these streams to the downstream community. Multivariate statistical methods were used to assess the relative importance of land use, local habitat conditions, and potential colonists to the downstream community.
Our mixing model used the relative abundances of taxa in the upstream reaches to predict the relative contribution of each of these reaches to the downstream community. Downstream communities were most similar to the community in only one upstream reach at 7 of 13 sites. At these sites, the upstream reach with the greatest contribution to the downstream community also had a significantly higher discharge than the other upstream reach. Linear regression analysis confirmed a highly significant relationship between the best fit mixing coefficients and the ratio of discharge from each upstream reach. Principle components analysis found three primary components that related to agricultural land use, watershed size and discharge, and physical habitat characteristics. Although community composition below confluences generally reflected the total amount of agricultural land use upstream, sites with contrasting land use in the upstream reaches had higher taxa richness than those with uniform upstream land use. These results reinforce the role of local habitat characteristics in influencing the degree to which macroinvertebrate communities below confluences are influenced by conditions upstream.
For Objective 2, experiments were conducted in six stream sites in the Lake Champlain Basin, Vermont during August 2001. Three agricultural sites were stream reaches surrounded by dairy farms. Three streams surrounded by forested land cover served as reference reaches. One-liter plastic tubs of various configurations that contained rocks devoid of macroinvertebrates were placed in the streams. Tubs suspended off the bottom of the stream captured drifting macroinvertebrates only. Tubs with screen lids submerged in the substrate collected macroinvertebrates moving through the hyporheic zone. Tubs with emergence traps collected emerging adult insects, and floating trays collected ovipositing adults and their eggs. Replicate sets of tubs were removed at 2, 7, and 14-day intervals. Drift nets also were used to capture macroinvertebrates present in the water column between dusk and dawn. Hess samples, which measure the resident benthic community, were collected in conjunction with drift net samples to assess whether the drifting community is indicative of the resident community.
These experiments showed rapid increases in density in all experimental treatments. Densities of colonists in the resident tubs were much higher at the agricultural sites than at the undisturbed sites by Day 7 of the experiment. Tolerant species, measured using the Hilsenhoff Biotic Index (HBI), were more abundant in agricultural streams than in undisturbed streams in both the drift tubs and the hyporheos tubs. The HBI was greater in resident tubs than Hess samples for all collection days, suggesting that although densities were similar, the resident tubs had not reached a stable community composition at the end of the 14-day experiment.
To address Objective 3, we conducted manipulative field experiments during June and July 2001 using flumes placed into one of our reference stream reaches. We used these flumes as experimental units to examine the effects of adding fine sediment, phosphorus, and their combination on stream macroinvertebrate and periphyton communities. We chose these two pollutants because they are well known byproducts of agricultural land use. Although the effects of each individually is relatively well known, the combined effects of these pollutants has not been investigated with dosages that are realistic for streams in northeastern watersheds.
We constructed 16 flumes within a 100 m reach of the Huntington River near Hanksville, VT. Each flume randomly was assigned one of four treatments: unmanipulated control, sediment addition, phosphorus addition, and sediment and phosphorus addition. In each flume, a set of macroinvertebrate colonization tubs was buried flush with the stream bottom and allowed to be colonized for 2 weeks prior the beginning of the experiment. On each of four treatment days, we applied fine and coarse sediment and soluble phosphorus for 12 hours. Water samples were collected for total phosphorus analysis before, during, and after treatment. In addition, we sampled for drifting macroinvertebrates at the start of the treatment day to assess immediate effects of the sediment treatment.
The experiment ran from June 21 (Day 0) to July 20 (Day 30). Treatments were applied on Day 1, Day 7, Day 19, and Day 23. A strong storm on July 1 and 2 (Days 10 and 11 of the experiment) created a very large spate that damaged some of our flumes. Repairs were conducted on July 5 and 6, 2001.
Samples of macroinvertebrates, periphyton, and sediments were collected on Day 1, Day 3, and Day 30. In each flume, three subsamples of macroinvertebrates and one sample of sediments were collected by removing four colonization tubs. Around each macroinvertebrate colonization tub, three rocks were scraped to collect periphyton for species enumeration and chlorophyll a analysis. Additional rocks were sampled from each flume for measurements of photosynthesis rates.
Unfortunately, the very large natural spate that occurred midway through our experiment and our problem with contamination of downstream flumes with phosphorus from upstream treatments reduced the value of this experiment. However, the data we obtained clearly show that periphyton species richness was reduced by the phosphorus treatment in our most upstream, uncontaminated flume. The phosphorus enrichment inside the flume enhanced photosynthesis and increased the abundance of Cocconeis sp. growing on the inside of the flume wall. This result is not surprising and is consistent with a hypothesis of phosphorus enrichment in the streams.
Macroinvertebrate drift significantly increased in those flumes that received sediment additions. The chironomids, in particular, responded to this disturbance by moving into the drift.
Conclusions:
Many studies of the edge-of-field effects of agricultural land uses on stream biota have documented significant changes at the local level, but to predict the implications at the watershed level, additional information about the interaction of land-use and stream architecture is needed. To move from impact assessment to effective siting and design of stream restoration, information about colonization patterns and recovery rates also is needed. In addition to restoring appropriate water quality and habitat conditions, successful stream restoration will require that there are sources of appropriate colonists upstream from potential restoration sites.
Similar to other studies, our work documents increased densities of stream
periphyton and macroinvertebrates at agricultural sites and a shift in species
composition towards pollution tolerant taxa. Our work also shows that downstream
communities constantly are receiving potential colonists from upstream reaches.
The diversity and abundance of the upstream macroinvertebrate colonists moving
both as drift and through the hyporheos also is influenced by land use, with
generally more movement at agricultural sites.
The high rates of movement and rapid colonization of new substrates by macroinvertebrates
suggests that when restoration provides appropriate substrate, recovery of stream
communities may be rapid. Using a simple mixing model, we found that we could
predict the species composition expected in downstream reaches based on the
relative abundances of taxa in the upstream reaches. Because discharge controls
the delivery of drifting colonists, it explained a large percentage of the variation
in the contribution of the upstream reaches. Those upstream reaches with a larger
discharge contributed proportionally more to the composition of the downstream
community.
Land use of the upstream reaches also was important in predicting downstream community composition. In general, community composition below confluences reflects the total amount of agricultural land use upstream. However, sites with contrasting land use in upstream reaches had higher taxa richness than sites with uniformly undisturbed or uniformly agricultural land use in the upstream drainage area. Because taxa richness only varied slightly between sites with different land use, these results suggest that local habitat characteristics may influence the degree to which macroinvertebrate communities below confluences are influenced by conditions upstream. These characteristics include pebble size, embeddedness, and shading.
Taken together, our results support a watershed approach to stream restoration. They suggest that restoration sites selected based on a full understanding of upstream land use and local habitat conditions will have the highest probability of success. If local habitat and water quality conditions are remediated, recovery of the biota should be rapid. Species composition might be predicted based on estimates of discharge and relative abundance of the upstream source populations.
Journal Articles:
No journal articles submitted with this report: View all 2 publications for this projectSupplemental Keywords:
agricultural pollution, ecological effects, biomonitoring, surface water, northeast., RFA, Scientific Discipline, ECOSYSTEMS, Water, Ecosystem Protection/Environmental Exposure & Risk, Water & Watershed, Ecosystem/Assessment/Indicators, Ecosystem Protection, exploratory research environmental biology, Chemical Mixtures - Environmental Exposure & Risk, Ecological Effects - Environmental Exposure & Risk, Ecological Monitoring, Terrestrial Ecosystems, Environmental Monitoring, Ecological Effects - Human Health, Ecological Risk Assessment, Ecological Indicators, Watersheds, nutrient dynamics, risk assessment, anthropogenic processes, ecosystem assessment, nitrogen deposition, forest ecosystems, sediment transport, agricultural watershed, Vermont, climate change, ecological assessment, ecosystem indicators, EPSCOR, aquatic ecosystems, water quality, lake ecosysyems, watershed assessment, ecosystem stress, climate variability, aquatic biota, ecological research, land use, restoration planning, watershed restorationProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.