From: Taxonomic macroinvertebrate sample identified to order and family level by volunteer or professional taxonomist.
To: DNA Barcoding used to identify macroinvertebrate sample identified to species level. Reveals the additional taxonomic resolution provided by DNA Barcoding.
RESULTS OF METHOD SELECTION
Four molecular analysis methods were evaluated, and had very different results in terms of the percent of samples that successfully amplified.
Silica resin and PCR bead successfully amplified 100% of the samples.
MACROINVERTEBRATE FAMILIES COMPARED: CHIRONOMIDAE CHOSEN
The Chironomidae samples showed the least undetermined nucleotides, best peak quality, and best Phred sequence quality.
The Gammaridae also responded very well to barcoding, However, the gammaridae do not have the range of geography and biotic indices that the Chironomidae do.
Phred scores were compared using two-sample t-tests (0.01 significance level).
Chironomidae vs. Physidae p = 1.01 x 10-6 indicating a statistically significant difference.
Chironomidae vs. Haliplidae p = 7.37 x 10-8 indicating a statistically significant difference.
Chironomidae vs. Gammaridae p = .053 indicating a difference that is not significant. However Gammaridae were not chosen due to their more limited number of species, geographic range, and biotic index range.
CHIRONOMIDAE HAVE A SURFACE GEOLOGY PREFERENCE
The Chironomidae sampled here aligned by genera with either high gradient streams in piedmont geology, or sandy soils and coastal plain geology. Only 13% of the genera sampled were found evenly in both geologies.
Coastal Plain bottom composition: silt
Piedmont bottom composition: cobble, pebble, bedrock
DISTRIBUTION BY NUTRIENT POLLUTION
A moderate positive linear association was found between the weighted average Hilsenhoff scores of the Chironomidae genera sampled at each site and nutrient pollution.
Comparison of Genera Tolerance Values and Nutrient Pollution: This analysis compared the relationship between the weighted average Hilsenhoff tolerance scores of the Chironomidae genera sampled at each site, with the nutrient pollution. The value for nutrient pollution was calculated from the average ppm of nitrate and orthophosphate sampled at each site, which was normalized to a value between 0 and 10. When nutrient pollution data for sites are graphed with the weighted average Hilsenhoff tolerance scores of the Chironomidae genera sampled, a moderate positive linear association is noted. There is a statistically significant relationship with p<0.05. In a linear regression ran, R2 =.67 indicating that 67% of the variation in the Hilsenhoff tolerance scores of the Chironomidae genera sampled were accounted for by overall nutrient pollution data. This means that 33% of the variation in tolerance score is influenced by factors other than nutrient pollution.
CHIRONOMIDAE DNA DATA SHOW A STRONG POSITIVE LINEAR RELATIONSHIP TO HISTORICAL HEALTH DATA
A strong positive linear association was found between the weighted average tolerance values of the Chironomidae genera sampled at each site, and the overall historical health based on sampling at each site.
Comparison of Genera Tolerance Values and Overall Historical Health Values: The Chironomidae health data correlates to historical health measurements. When historical health data for sites are graphed with the weighted average Hilsenhoff tolerance scores of the Chironomidae genera sampled, a strong positive linear relationship is noted. There is a statistically significant relationship with p<0.05. In a linear regression, R2=.79 indicating that 79% of the variation in the Hilsenhoff tolerance scores of the Chironomidae genera sampled were accounted for by overall historical waterway health data. This means that 21% of the variation in tolerance score is influenced by factors other than overall waterway health. Bottom composition is likely a part of that 21% as there are some Chironomidae genera that prefer a healthy pebble-bottomed stream over a healthier mud-bottomed stream.
CHIRONOMIDAE DNA DATA COMPARE FAVORABLY TO CURRENT MANUAL METHOD
A Bland-Altman analysis showed limits of agreement of -0.853 and 0.868 between the weighted average tolerance values of the Chironomidae genera barcoded and the standard method that uses manual taxonomic identification by morphology.
This indicates that the new method proposed here of DNA barcoding Chironomidae is in agreement with the current method to within 1.72 on a zero to ten health
scale.
RESULTS OF PHYLOGENETIC TREE ANALYSIS
The following phylogenetic trees were used to analyze the genetic relationships between selections of the
Chironomidae sampled with respect to site, taxa level identified, and biotic index.
Phylogenetic tree which diagrams the genetic relationship between the Chironomidae samples from six sites with the most variation:
Phylogenetic tree which diagrams the genetic relationship between Chironomidae samples with the taxa level identified (e.g, subfamily, genus, species):
Identification down to species level indicates a match in the sequence databases. Identification to genus or subfamily indicates gaps in the sequence database that can be filled with a widespread barcoding initiative. The gaps could also allude to potential novel species.
Phylogenetic tree which diagrams the genetic relationship between Chironomidae samples with the Hilsenhoff tolerance value for each genera:
The family biotic index for Chironomidae is 6. This masks an underlying variability as the genera sampled for this study range in biotic index from 2 to 10 on a scale of 0 to 10 health scale.