Research Topics > Ecology > printer-friendly format

Effect of Reach-Level Changes in Ditch Morphology
and Riparian Vegetation on Macroinvertebrate Communities
in an Agricultural Landscape

Gregg Sablak
Environmental Science Graduate Program

Virginie Bouchard, Professor
Environmental Science Graduate Program
and School of Natural Resources

The Ohio State University
Columbus, Ohio

Introduction

This research is a part of a large effort to improve stream ecosystem integrity and water quality within agricultural landscapes, while maintaining drainage function. Two-stage ditches that include a low depositional floodplain are being proposed as a viable alternative to the current one-stage ditch design. This feature is expected to increase channel stability and create habitat. Our objective was to determine the relative effects of riparian tree development and two-stage channel morphology on the benthic macroinvertebrate community.

Our work addresses important aspects of channel restoration within heavily farmed areas, which is highly relevant to Ohio and the Midwestern United States.

• Crop agriculture accounts for 24% land use in the lower 48 states
• Agriculture is listed by US EPA as the #1 cause for water quality impairment in rivers and streams
• Channel modification due to drainage improvement efforts is a significant factor for this impairment
• Agriculture is indispensable and soil drainage is important for productivity

Addressing Stream Impairment

• Riparian buffers are used to mitigate nonpoint source pollution
• Both grassed and forested buffers are shown to be effective
• Proposed strategies for improving drainage ditch design has strong implication for riparian zone
• Two stage ditch design

Stream Ecology

• Strong link between riparian zone and benthic community in small streams
• Disturbance may be a primary factor shaping stream communities
• Aquatic community responds to land use conditions at local (reach level) scales

One-Stage Ditches

One-stage ditch

One-stage ditch at high water

Newly constructed one-stage ditch

Charachteristics of one-stage ditches:

• Unstable morphology leading to extreme erosion/deposition processes
• Intensified storm flow velocity and shear stress
• Over-widened lower energy base flow channel
• Little flow heterogeneity
• Typically fine bed substrate (poor quality)
• Sparse development of habitat and flow refugia

Two-stage ditches

Two-stage ditch

Two-stage ditch at high water

Newly constructed two-stage ditch

Charachteristics of two-stage ditches:

• Successional to the one-stage ditch, and formed by erosion and deposition processes
• May not develop in presence of riparian trees
• Appears to be more stable form
• May provide reduction in storm flow velocity and shear stress
• Narrower higher energy base flow channel
• Better riparian connection
• Potentially improves habitat and flow refugia

Tree-Lined Ditches

Tree-lined ditch

Tree-lined ditch at high water

Outside view of a tree-lined ditch

Charachteristics of tree-lined ditches:

• Often has one-stage morphology
• Favor over-widened, lower energy base flow channel
• Intensified storm flow velocity and shear stress
• Provides debris, detritus, and structure that can be used as habitat, refugia and food resources
• Reduces incident sunlight, and maintains lower temperatures

Objective and Hypotheses

Determine the relative effects of ditch morphology and riparian vegetation on benthic community assemblage at the reach level.

H1: Trees provide better habitat and food diversity than open canopied sites which enhances the benthic community.

H2: Small floodplain benches within ditches reduce the severity of disturbance events and improves habitat, which enhances the benthic community.

Methods

General Design

• Study reaches (150m long) selected a-priori
  
  
• Control (n=6): No mature trees; bw/cw < 0.3
  
  
• Bench (n=6): No mature trees; bw/cw > 0.5
  
  
• Tree (n=7): Mature trees on both banks

Field Work

Benthic Community (sampled June 15 to July 15)

• 6-10 Hess samples per reach
  
  
• Each reach sampled one time in either 2001 or 2002

Morphology and Substrate (surveyed spring, summer or fall)

• 2-4 cross-sections and longitudinal profile
  
  
• Modified Wolman Pebble Count
  
  
• Performed in either 2001 or 2002

Habitat Evaluation

Qualitative Habitat Evaluation Index (QHEI, Ohio EPA) completed for each site

Six category separations used:

1. Substrate
2. In-stream cover
3. Channel morphology
4. Riparian conditions
5. Riffle/run development
6. Pool/glide development

Laboratory Work

• Separated macroinvertebrates from sample debris by visual inspection
  
  
• Identified to lowest possible taxon (genus or species) and enumerated macroinvertebrates
  Exceptions:
  • Chironomidae (midges)
  • Oligochaeta (aquatic worms)
  • Hydracarina (water mites).

Computations and Data Analysis

• Morphology Metrics
  • Bench development, water depths, channel gradient, wetted perimeter
    
    
• Substrate Metrics
  • Texture class composition (relative %) of bed substrate
  • “Percent finer than” parameters (D16, D35, D50, D84, D95)
    
    
• Macroinvertebrate Community Metrics
  • Taxa richness
  • Shannon diversity
  • Density
  • %EPT
  • Modified ICI
  • Rarefaction
    
    
• ANOVA and nonparametrics (Kruskal-Wallis and Mann-Whitney) analyses used for group comparisons
  
  
• Forward and backward stepwise linear regressions

Results

Findings

Bench Development

Bench Width to Channel Width Ratio: Bench width to channel width ratio for all study reaches. The mean for the Bench group (1.04 +/- SD 0.51) is significantly higher than both the Tree (0.26 +/- SD 0.14) and Control (0.20 +/- SD 0.08) groups.

Substrate Quality

Texture Class Composition of Bed Substrate: Mean relative percent composition of texture classes for each group. The % gravel is significantly higher in the Bench group than in both the Control and Tree groups. The % silt/clay is significantly lower in the Bench group than in the Control and Tree groups.

Habitat Quality

OEPA: Qualitative Habitat Evaluation Index (QHEI) Overall Score: Box plots of QHEI total score values for each group. Red dots indicate mean values, the center line is the median, the tops and bottoms of the boxes are the upper and lower quartiles respectively, and the whiskers represent the extent of the data. Both the Bench and Tree groups are significantly greater than the Control group.

Taxonomic Richness

Rarefaction Curves: Rarefaction curves for each group based on pooled samples. Curves have not been analyzed for statistical differences. Trend shows small differences in the taxonomic richness per number of individuals collected. The tree group is highest, followed by the Bench and then the Control groups.

Macroinvertebrate Metrics: Means and standard deviations of macroinvertebrate community metrics for each group. No differences are statistically significant. The % EPT (Ephemeroptera, Plecoptera and Trichoptera) is the relative composition of mayfly, stonefly, and caddis fly taxa. These taxa are typically pollution sensitive, and therefore are indicators of the quality of streams.

Community Assemblage

Feeding Guild Distribution: Mean feeding guild distribution across the three groups based on pooled samples. There were no statistically significant differences across the groups for any of the feeding guilds. Members of the Chronomidae (midges) were evenly distributed in the collector-gatherer and predator guilds.

Macro Invertebrate Community Index Scores: Box plots of modified Invertebrate Community Index (ICI) values for each group. Red dots indicate mean values, the center line is the median, the tops and bottoms of the boxes are the upper and lower quartiles respectively, and the whiskers represent the extent of the data. Both the Tree and Bench groups are significantly greater than the Control group.

Stepwise Linear Regression

Response and predictor variables and correlation coefficients for all significant forward and backward stepwise linear regressions. Macroinvertebrate metrics were related to morphology, substrate, and habitat. Habitat and substrate texture were positively influenced primarily by two-stage morphology. The (BE), or bench elevation, is the elevation in which the morphology metric was derived (i.e., the bench is inundated) . The other elevation used in deriving morphology metrics was the top of the channel, which is slightly lower than the bench elevation (i.e., the bench is not inundated).

Conclusions

• Bench development provided the best substrate quality and habitat for these ditches.
  
  
• Benthic communities had a marginal response to reach scale changes in riparian vegetation and ditch morphology.
  
  
• Density of individuals was weakly related to the amount of cobble in the substrate.
  
  
• The presence of either trees or benches resulted in an increase in relative composition of sensitive taxa.
  
  
• Feeding guild structure was not influenced by riparian vegetation or channel morphology.
  
  
• Taxa richness may not strongly reflect changes in benthic communities relative to riparian vegetation and channel morphology in heavily farmed areas.

Acknowledgements:
This research was funded by the Great Lakes Protection Fund.

Special Thanks to:

• OEPA Ecological Assessment Unit
• Dan Mecklenburg (ODNR)
• Ed Rankin (ILGARD)
• Andy Ward and Larry Brown (OSU, FABE)