Yellow River IA 01-YEL-436

[Note: A TMDL for the bacterial impairment for streams in the Yellow River basin was prepared by Iowa DNR in 2012 and was approved by U.S. EPA in February 2013. Due to completion of the TMDL, the Integrated Report categories for the Class A1 and Class A2 recreational uses are changed from 5p (impaired; TMDL needed) to 4a (impaired; TMDL approved). Because not all impairments identified for this assessment segment are covered by the TMDL (aquatic life impairment due to low dissolved oxygen), this segment remains in Category 5a of Iowa's Integrated Report (impaired and TMDL required).]

Both the Class A1 (primary contact recreation) and Class A2 (secondary contact recreation) uses remain assessed (evaluated) as “not supported” due to violations of Iowa’s water quality criteria for indicator bacteria. The geometric mean of E. coli in the 13 samples collected at DNR/SHL station 11030001 during the recreational season of 2009 was 13,288 orgs/100 ml. This geometric mean far exceeds the Class A1 criterion of 126 orgs/100 ml. All 13 samples (100%) exceeded Iowa’s Class A1 single-sample maximum criterion of 235 orgs/100 ml. This geometric mean also far exceeded the Class A2 criterion of 630 orgs/100 ml. Nine of the 13 samples (69%) exceeded Iowa’s Class A2 single-sample maximum criterion (2,880 orgs/100 ml.). According to U.S. EPA guidelines for Section 305(b) reporting and DNR’s assessment/listing methodology, if the geometric mean of E. coli is greater than the applicable state criterion, the primary contact recreation uses should be assessed as "impaired" (see pgs 3-33 to 3-35of U.S. EPA 1997b). Thus, because the recreation season geometric mean for DNR/SHL station 11030001 exceeded both the Class A1 and Class A2 criteria, these uses were assessed as “not supported.”

A fish kill occurred on this stream reach on March 17, 2000. An estimated 4,800 fish were killed; no cause or source of the kill was identified. The following account is from the DNR fish kill database and includes comments from the DNR staff that investigated the kill:

"Dead fish were observed both upstream and downstream of the Smith Road bridge on the Yellow River, but only below the confluence with the unnamed tributary in section 17. North of Postville: water turbid and green, with high flow. The kill affected a 3.1 mile segment of the Yellow River. Dead fish were not observed in the unnamed trib that enters into the Yellow River in Sec. 17. A total of 12 species of fish were affected. After visually inspecting the habitat in the area that the fish kill occurred, we would rate the habitat as excellent. The bottom substrate of much of the stream is composed of gravel and cobble. Deeper water areas for fish overwintering locations are also available. The Smith Road Bridge is the upstream extent of the area in which we have stocked fingerling brown and rainbow trout. No trout were seen in any of the counts of dead fish. The total number of fish that were observed was very low for a stream of this size and habitat quality. In previous work by Paragamian (1986), between 3,000 and 5,000 fish/acre were seen in northeast Iowa streams of comparable sizes. We feel that there has been an unknown factor at work that has severely reduced the total fish community in this section of the Yellow River prior to the total kill that this incident has caused to occur."

In recent IR cycles (i.e., 2004-2012), the assessment of the impact of the 2000 fish kill on this stream segment has been considered a “potential impairment” and thus placed in non-303(d) IR Category (IR 3b) indicating that additional monitoring and investigation was needed. Based, however, on a review of 2000 fish kill report and the rationale for moving this impairment to IR Category 3b, the 2000 fish kill was moved to IR Category 5b for the following reasons: (1) the relatively large number of fish killed (~4,800), (2) the length of the kill (3.1 miles), and (3) the history of chemical/physical water quality issues in the segment (e.g., high violation frequency of dissolved oxygen criteria and extremely high E. coli levels). Based on this information and on other comments from DNR staff that investigated the kill (see above), some type of pollutant cause is suspected of having caused this kill. Follow-up monitoring is needed to determine the status of the fish community in this assessment segment. Note: the movement of the March 2000 fish kill impairment from IR 3b (potentially impaired) to IR 5b (impaired) was to have taken place during the 2014 IR cycle: this move did not occur. Thus, this impairment was moved to IR 5b for the 2016 cycle.

The assessment of the Class B(CW1) aquatic life uses was also based on data collected in 2011 and 2012 as part of the DNR/SHL coldwater stream sampling projects. A series of biological metrics which reflect coldwater stream water quality and habitat integrity were calculated from the biological sampling data. The biological metrics are based on the numbers and types of benthic macroinvertebrate taxa that were collected in the stream sampling reach. The biological metrics were combined to make a coldwater benthic index (CBI). The index ranks the biological integrity of a coldwater stream sampling reach on a rising scale from 0 (minimum) to 100 (maximum). The 2011 CBI score was 60. The 2012 CBI scores were 59.4 and 50. The aquatic life use support was assessed (monitored) as fully supporting (=FS), based on a comparison of the CBI scores with biological impairment criteria (BIC) established for the 2012 Section 305(b) report. The biological impairment criteria were determined from a statistical analysis of data collected at coldwater stream reference sites from 1994-2011. The CW BIC is 60 and this segment passed the CW BIC 1/3 times in a four-year period (2011-2014). DNR normally requires a segment to have a simple majority of passing scores from multiple years in a five-year period to be considered fully supporting (monitored). However, this assessment of this segment of the Yellow River includes the professional judgment of DNR staff incorporating an in-depth analysis of the metric scores, sampling conditions and the fish sampling that occurred at the time the benthic macroinvertebrates were sampled. A fish sample from 2012 contained over 5000 fish per mile and included 337 Brown Trout of various sizes and one Rainbow Trout.

Fish consumption uses remain “not assessed” due to the lack of fish contaminant monitoring in this stream segment.