Recurrent die-offs of adult coho salmon returning to spawn in Puget Sound lowland urban streams

Abstract

Several Seattle-area streams in Puget Sound were the focus of habitat restoration projects in the 1990s. Post-project effectiveness monitoring surveys revealed anomalous behaviors among adult coho salmon returning to spawn in restored reaches. These included erratic surface swimming, gaping, fin splaying, and loss of orientation and equilibrium. Affected fish died within hours, and female carcasses generally showed high rates (>90%) of egg retention. Beginning in the fall of 2002, systematic spawner surveys were conducted to 1) assess the severity of the adult die-offs, 2) compare spawner mortality in urban vs. non-urban streams, and 3) identify water quality and spawner condition factors that might be associated with the recurrent fish kills. The forensic investigation focused on conventional water quality parameters (e.g., dissolved oxygen, temperature, ammonia), fish condition, pathogen exposure and disease status, and exposures to metals, polycyclic aromatic hydrocarbons, and current use pesticides. Daily surveys of a representative urban stream (Longfellow Creek) from 2002-2009 revealed premature spawner mortality rates that ranged from 60-100% of each fall run. The comparable rate in a non-urban stream was <1% (Fortson Creek, surveyed in 2002). Conventional water quality, pesticide exposure, disease, and spawner condition showed no relationship to the syndrome. Coho salmon did show evidence of exposure to metals and petroleum hydrocarbons, both of which commonly originate from motor vehicles in urban landscapes. The weight of evidence suggests that freshwater-transitional coho are particularly vulnerable to an as-yet unidentified toxic contaminant (or contaminant mixture) in urban runoff. Stormwater may therefore place important constraints on efforts to conserve and recover coho populations in urban and urbanizing watersheds throughout the western United States.

Figure 1. Stream survey and sample collection locations.

The map indicates the greater Seattle metropolitan area, with gray shading representing the relative intensity of urbanization. Shown are the urban and non-urban creeks surveyed for coho spawner mortality, regional hatcheries, and the location of seawater-phase adult coho collections along the Seattle waterfront (Elliott Bay).

Figure 2. Representative adult female coho carcass with characteristically high egg retention.

Rates of premature spawner mortality within and across urban drainages were quantified on the basis of egg-retaining female carcasses. Shown is a female affected by the mortality syndrome in Longfellow Creek in the fall of 2005. In most cases, egg retention was nearly 100%. Photo by Tiffany Linbo, NOAA Fisheries.

Figure 3. Prematurely dying spawners do not show evidence of neurotoxic pesticide exposure.

Shown are relative rates of brain acetylcholinesterase (AChE) activity, a target enzyme for common homeowner use insecticides, in adult coho salmon. The brain enzyme activities of affected fish from an urban stream (Longfellow Creek; LF) were not significantly inhibited relative to unaffected fish from a non-urban stream (Fortson Creek; FT) and three regional hatcheries; Issaquah (ISS-H), University of Washington (UW-H), and Stillaguamish (STI-H). Error bars are 1 standard error of the mean. Sample size is indicated in parentheses and letters indicate significant differences between locations (one-way ANOVA, Tukey's HSD; p<0.05).

Figure 4. The gill tissues of prematurely dying coho contain elevated levels of cadmium, lead, and nickel.

The concentrations of metals in the gills of affected coho from two urban streams (Longfellow and Des Moines Creeks) are plotted relative to samples collected from a non-urban hatchery (Wallace River). Error bars are 1 standard error of the mean. Letters indicate significant differences in measured levels of metals between sampling locations (one-way ANOVA, Tukey's HSD; p<0.05) and horizontal bars indicate no significant differences (p>0.05).

Figure 5. Analysis of bile from affected coho spawners indicates exposure to polycyclic aromatic hydrocarbons (PAHs).

Concentrations of fluorescent PAH metabolites (as phenanthrene [PHN] and benzo-a-pyrene [BaP] equivalents) in the bile of coho collected in an urban stream (Longfellow Creek; LF) are shown relative to fish from a non-urban stream (Fortson Creek; FT) and a non-urban hatchery (Wallace River Hatchery; WR-H). In 2003, seawater-phase coho (pre-freshwater entry) were also sampled from urban Elliott Bay (EB). The bile data demonstrate a significant increase in PAH exposure after coho spawners transition from a highly urbanized estuary to freshwater spawning habitats. Sample sizes, the same for PHN and BaP, are indicated over each bar. Error bars are 1 standard error of the mean.

Figure 6. Pre-spawn mortality and survival to spawn in relation to rainfall.

Shown are the results of daily stream surveys throughout the 2002 coho spawning season in Longfellow Creek in relation to daily rainfall. Asterisks (*) indicate days when stream flows were too high to survey the creek.