EPA's Vessel General Permit and Small Vessel General

other tributaries flowing into Ozette Lake, Washington. Two artificial propagation programs are
considered part of this ESU: The Umbrella Creek and Big River sockeye salmon hatchery
programs. NMFS considers these artificially propagated populations no more divergent relative
to the local natural population than would be expected between closely related natural
populations (70 FR 37160).
The historical abundance of Ozette Lake sockeye salmon is poorly documented, but may have
been as high as 50,000 individuals (Blum 1988). The overall abundance of naturally–produced
Ozette Lake sockeye salmon is believed to have declined substantially from historical levels. In
the first study of lake escapement of Ozette Lake sockeye salmon (Kemmerich 1945), the run
size entering the lake was estimated at a level of several thousand fish. These counts appear to
be roughly double the current mean lake abundance, considering that they were likely conducted
upstream from fisheries in or near to the Ozette River. Makah Fisheries Management (as cited in
Good et al.,, 2005) concluded that there appears to be a substantial decline in the Tribal catch of
Ozette Lake sockeye salmon beginning in the 1950s and a similar decline in the run size since
the 1920s weir counts reported by Kemmerich (1945).
An analysis of total annual Ozette Lake sockeye salmon abundance (based on adult run size data
presented in Jacobs et al., 1996) indicates a trend in abundance averaging -2% per year over the
period 1977 through 1998 (NMFS 1998d). The current tributary-based hatchery program was
planned and initiated in response to the declining population trend identified for the Ozette Lake
sockeye salmon population. The most recent (1996 to 2003) run-size estimates range from a low
of 1,609 in 1997 to a high of 5,075 in 2003, averaging approximately 3,600 sockeye per year
(NMFS 2009). For return years 2000 to 2003, the 4-year average abundance estimate was
slightly over 4,600 sockeye. Because run-size estimates before 1998 are likely to be even more
unreliable than recent counts, and new counting technology has resulted in an increase in
estimated run sizes, no statistical estimation of trends is reported. The current trends in
abundance are unknown for the beach spawning aggregations. Although overall abundance
appears to have declined from historical levels, whether this resulted in fewer spawning
aggregations, lower abundances at each aggregation, or both, is not known (Good et al. 2005).
Based on estimates of habitat carrying capacity, a viable sockeye salmon population in Lake
Ozette watershed would range between 35,500 to 121,000 spawners (Rawson et al. 2009 as cited
in NMFS 2009).
There has been no harvest of Ozette Lake sockeye salmon for the past four brood-cycle years
(since 1982). Prior to that time, ceremonial and subsistence harvests by the Makah Tribe were
low, ranging from 0 to 84 fish per year. Harvest has not been an important mortality factor for
the population in over 35 years. In addition, due to the early river entry timing of returning
Ozette Lake sockeye salmon (beginning in late April, with the peak returns prior to late-May to
mid-June), the fish are not intercepted in Canadian and U.S. marine area fisheries directed at
Fraser River sockeye salmon. There are currently no known marine area harvest impacts on
Ozette Lake sockeye salmon.
Overall abundance is substantially below historical levels (Good et al. 2005). Declines in
abundance have been attributed to a combination of introduced species, predation, loss of
100