Location of ODFW adult and juvenile salmon traps and the coho spawning reach in the Winchester Cr system. Coastal Frontal Streams indicated in red are the locations of (north to south) Big, Five Mile, Three Mile and Two Mile Creeks for which ODFW has quantitative and qualitative salmon presence data.

Evaluation: Significant Action Needed/Closely Monitor

Issue Summary:

The Winchester Creek coho salmon run is in a vulnerable state. There is only one spawning reach in the Winchester system and marine survival rates are highly variable but generally lower than other systems. Runs that existed historically in the coastal frontal streams (other than Big Creek) are considered at or near zero.

Why do we care:

Salmon are important species for our area in terms of their commercial and recreational value, iconic cultural status, and as indicators of coastal ecosystem health.

What’s Happening

The spatial extent of salmonids use of streams in the South Slough and Coastal Frontal watersheds is mapped in Figure 1. Overall, most streams are considered to be populated with coastal cutthroat trout. Coho salmon are considered to be present in most South Slough tributaries but only in the Coastal Frontal area’s Big Creek system. Coho spawning occurs only in very limited areas (see below). Winter steelhead also use most the South Slough tributarues and Big, Two Mile, Three Mile and Five Mile Creeks in the Coastal Frontal area. There is limited use of the lower South Slough estuary by fall Chinook.

ODFW has also identified actual or potential fish barriers which include problem culverts, natural water falls and an unspecified type of dam (probably a beaver dam) (Figure 1).

Yearly adult and juvenile coho salmon populations in South Slough’s Winchester Creek have been tracked by the Oregon Department of Fish and Wildlife’s (ODFW) life cycle monitoring program since 2000.

ODFW’s Bruce Miller (retired) analyzed 11 years of those data which suggest that marine survival for coho salmon in South Slough’s Winchester Cr is potentially in trouble. The average marine survival rate over the years we have data is 3.16% with a relatively low standard error which indicates there’s not huge variation in that percentage year to year (Table 1). The greatest percentage of returns in Winchester Creek was 8.56% in 2001. The smallest return percentage was 0.06% in 2007. There is no real trend to the overall 11 year dataset (Figure 1).

These smolt to adult return rates are low when compared to other coho salmon runs also tracked by ODFW. For example, average percent coho salmon marine survival is 7.23% in the West Fork Smith River, with the range of marine survival rate being over 20% in 2003 and under 1% in 1999 (Table 1 and Figure 2). There is no real trend to the overall 12 year dataset for West Fork Smith River returns. Marine survival rates for coho salmon were also better in two other estuaries, Nehalem and Siletz (Figure 3). It is worth noting that there may to be a pattern in coho salmon marine survival rates in Winchester Creek, Smith River, Nehalem, and Siletz populations. Survival rates between 2000 and 2004 may generally have been greater than survival rates between 2005 and 2008 in all four estuaries (Figure 3).In the coastal frontal area, there are not a lot of data available. Bruce Miller compiled available data from Big Creek, Five Mile, Three Mile and Two Mile Creeks.

Overall it appears that of the four major Coastal Frontal streams for which we have information, Big Creek, Five Mile Creek, Three Mile Creek, and Two Mile Creek only Big Creek continues to support a notable population of coho salmon. The other three streams are not currently major salmon producers. Although it’s not clear how many salmon they produced in the past Three Mile creek is known to have had coho runs. It’s likely that Five Mile and Two-Mile Creeks had coho runs too.

According to recent ODFW salmon spawning surveys in two Big Creek tributaries, there appears to be some limited spawning occurring in one tributary but very little in the other with more spawning occurring during more recent years than during past years (Table 2 and Figure 4).

Most of the Big Creek tributaries are designated as fish streams according to Oregon Department of Forestry’s fish presence data (Figure 4). Likewise, the majority of the Five Mile, Three Mile and Two Mile Creek tributaries are also designated as fish streams (Figure 5).

Anecdotal information about Five Mile, Three Mile and Two Mile Creeks was compiled by Bruce Miller from stream habitat, spawning, and juvenile fish surveys conducted between 1969 and 2010 (Table 3). Overall it appears that these creeks have not supported coho salmon populations for some time but cutthroat populations (presumably resident?) were “widely distributed” in Five Mile Creek during 1969, 1995 and 2010 surveys; “present” in Three Mile Creek during 1969 and 2010 surveys; and “present” in Two Mile Creek during 1969, 2005 and 2010 surveys. Stream habitat in all three creeks appears to have been affected by historic land use practices. ODFW Aquatic Inventories Project: Fish Survey Data


Five Mile Creek:   Surveys Results
Stream habitat   survey (1969) Predominantly   fine in- stream sediments, limited spawning gravel, adjacent hillsides   clear-cut, no fish-passage barriers
Spawning surveys   (1992, 1993, 1994) No coho salmon   observed
Juvenile fish   surveys (1969-seine, 1995 and 2010- electro-fisher) No coho salmon   observed;  cutthroat trout widely   distributed
Three Mile Creek:   Surveys Results
Stream habitat   survey (1969) Predominantly   fine in- stream sediments, little spawning gravel, adjacent hillsides   clear-cut, no fish-passage barriers
Spawning surveys   (none, officially) Unofficial note   (1969):  “local residents noted salmon   spawned in creek before opening of Seven Devils Mine” [chromite mine,   operated 1943]
Juvenile fish   surveys (1969-seine, 2010- electro-fisher) No coho salmon   observed, only cutthroat trout.  Partial   to total coho/steelhead  barrier at the   county road culvert just upstream from the ocean beach documented in   2010.  Total barrier to cutthroat at   most flows due to perch and velocity in pipe.
Two Mile Creek:   Surveys Results
Stream habitat   survey (1969- entire creek; 2005- 0.5 km reach) Predominantly   fine in-stream sediments, little spawning gravel, lower reach gorse-covered   pasture, upper reaches clear-cut, no fish-passage barriers
Juvenile fish surveys   (1969-seine, 2005 and 2010- electro-fisher) No coho salmon   observed, only cutthroat trout

Why Is It Happening?

Marine survival

For the Winchester Creek coho salmon run, marine survival may be higher than reported because there’s evidence to suggest that salmon returning to Winchester Creek are straying to other systems (based on a single year of detections of PIT tagged Winchester Creek smolts detected at PIT antennas in Palouse and Larson Creeks in the upper Coos estuary). Straying occurs in all systems but it may be particularly significant for Winchester Creek because it’s a relatively small stream flowing through South Slough at the mouth of the Coos River system. Returning fish may miss “smelling” Winchester Creek as they pass by the mouth of South Slough and head up the bay (Bruce Miller, pers. comm.). Additional information is needed to improve our understanding of Winchester Creek salmon straying. (predation: seals, sea lions, cormorants?)

Rearing habitat in the Winchester Creek system

A great deal of work has been done over the past 11 years to quantify the number of adult salmon returning to, spawning in, and juvenile salmon produced by the Winchester Creek system each year: ODFW coho life cycle monitoring program. In addition, South Slough NERR staff teamed with ODFW staff to characterize the behavior of juvenile salmonids in the South Slough estuary suggesting the key role estuarine rearing plays in marine survival for coho salmon (Miller and Sadro 2003).

The South Slough NERR has also reestablished structural complexity and physical and biological functions to formerly degraded salt marshes and freshwater wetlands in South Slough’s upper estuary through a series of habitat restoration and large woody debris placement projects (1996-2004; e.g., Cornu and Sadro 2002). One of the primary goals of this work was to improve juvenile salmonid habitat.

We have characterized those improvements with some electro-fishing, seine netting, acoustic tagging and underwater videography work done mainly in tidal and non-tidal channel reaches of Winchester Creek and Dalton, Cox, and Anderson Cr. marshes. In some cases the response was dramatic. For example, in Dalton Creek marsh tidal channel, a population of approximately 50 juvenile salmonids was present in the newly constructed tidal channel only a few months after construction. Mark-recapture methods revealed juvenile salmonid growth rates were greater than those for fish in Winchester Creek; electroshocking in the beaver dam constructed at the Cox Marsh restoration project two years after the restoration project was implemented, revealed significant populations of juvenile salmonids with the greatest growth rates of all the mark-recapture work done in South Slough (Miller and Sadro 2003).

In other cases the response was limited. For example, the pool habitat formed in Anderson Creek in the years after the restoration project was implemented were not sufficiently numerous or deep to accommodate juvenile coho salmon. Most of the pools were inhabited by the more aggressive and less habitat-sensitive juvenile cutthroat trout. This information led to the installation of large wood root wads that were placed in the creek to scour additional pools. Greater pool formation would lead to improved summer habitat for salmonids- meaning that if conditions permitted, juveniles could hold in Anderson creek all summer and not have to return further up the system to find cool freshwater refuge.

In general, although additional work can still be done to further improve estuarine conditions in the upper South Slough estuary (summer rearing habitat in particular), estuarine habitat should not be a major limiting factor to coho salmon in the Winchester system.

Spawning habitat

Spawning habitat is an entirely different story. There is currently a single, one km reach in the Winchester system (upper west fork Winchester Cr.) that supports sustained coho spawning. There is evidence that spawning occurs elsewhere in the system (probable redd sightings and presence of juvenile coho above juvenile barriers), but because the spawning habitat everywhere else in the Winchester system is so poor, those instances where coho spawning occurs outside the west fork are likely to be rare. However, anecdotal observations by ODFW staff over the past 10 years suggest spawning habitat has improved. One hypothesis offered for the improvement is the lack of recent sediment inputs and flushing of previous sediment loads (Bruce Miller, pers. comm.).

The county has plans in 2012 to harvest timber in the west fork adjacent to the lower part of the spawning reach and above it. The other forested slopes along that reach will likely be harvested within the next 2 to 5 years. So spawning habitat is very much a limiting factor for coho salmon in the Winchester Creek system.

Coastal Frontal stream systems

For the Coastal Frontal stream systems it would be easy to speculate that unlimited past land practices including timber harvest and associated road building, and mining with little or no awareness of the long-term effects of those activities could very well have changed salmon stream habitat.

Literature cited

Cornu, C. E., and S. Sadro. 2002. Physical and Functional Responses to Experimental Marsh Surface Elevation Manipulation in Coos Bay’s South Slough. Restoration Ecology 10: 474-486.

Miller, B., and S. Sadro. 2003. Residence Time and Seasonal Movements of Juvenile Coho Salmon in the Ecotone and Lower Estuary of Winchester Creek, South Slough, Oregon. Transactions of the American Fisheries Society 132: 546-559.