Idaho Sockeye Salmon and Global Warming: A Personal Perspective

Across much of the Pacific Northwest, the winter of 2014-2015 was quite dry and much of the area is still experiencing some level of drought. Except for coastal areas, this is a region of snowmelt dominated watersheds: After the snow melts, river flows fall dramatically and in drought conditions stable summer flows can be quite low indeed. Of course, this is also the home of tremendous populations of anadromous fish*, most notably various stocks of salmon and steelhead in the Columbia River. Because of the drought and warmer than normal water temperatures in the Columbia River system, there was much discussion over the effects of climate change on these fishes, most of which are listed as "Threatened" or even "Endangered" under the Endangered Species Act (ESA). One article in particular caught my attention as it was making its rounds in the Twitterverse because it was both of local interest and because I was asked what I thought about by someone I occasionally converse with on Twitter.

This article's byline was Boise, Idaho, and the species of interest was Sockeye Salmon (Oncorhynchus nerka). I'm not an expert on sockeye salmon or their habitat, but this is not a technical essay on sockeye per se, it's more of an opinion piece. But I do know that sockeye salmon have been nearly extinct in Idaho for decades, and have been considered functionally extinct. Presently, expensive captive broodstock and hatchery programs seem to be helping the Idaho's sockeye recover to a limited extent, but the landlocked form (kokanee) is most common in Idaho now.

I will start by stipulating that Columbia River and its tributaries experienced unusually warm water in 2015. I will also stipulate that water temperatures in the lower river have been gradually increasing; this is also true in Canada's Fraser River. A Seattle Times article in June claimed the river at Bonneville Dam (the farthest downstream) was hotter than it had been since 1950. In July, Snake River water temperature at Lower Granite Dam (the last passable major dam) was reported as 68°F** and few sockeye were passing through; some 50 were captured and transported by truck to the hatchery in Eagle, Idaho (see here). But is this situation a clear result of anthropogenic global warming (AGW)? Given that river temperatures may have been as high in the recent past, is this situation even unprecedented? Or could climate change, which has likely been occurring since the mid-1700s (even if exacerbated by human actions), be just one of many other better proved threats to anadromous fishes?

To start with, sockeye salmon were in a population decline in the 1950s or so, before the concept of AGW was conceived  and about the time adherents of climate change theory think anthropogenic CO2 emissions started to become an issue; this is some decades before AGW became widely accepted. Between 1937 and 1975, eight hydroelectric dams were built on the Columbia River and it's major tributary the Snake River, dramatically impairing most of Idaho's anadromous fish from reaching the Salmon River, one of the largest free flowing rivers in the continental United States. These dams mostly support some level of impaired fish passage, but many other parts of Idaho had their salmon runs extinguished as the Columbia River hydrosystem was developed***, sometimes by impressive structures like Hells Canyon Dam shown here.

But there are many full and partial barriers scattered throughout the system, and effects to anadromous fishes started before the first major hydroelectric dams were built. In fact, the ability of sockeye to acccess the upper Salmon River and the Sawtooth Valley was initially blocked by Sunbeam Dam in 1913, though this was probably not a complete barrier and was removed by 1934.

Sockeye (and kokanee) salmon have inhabited the Columbia River watershed since at least the Wisonsinian Glaciation about 14,000 years ago†. Sockeye were never really very abundant in Idaho compared with other salmon and with sockeye populations in Canada's Fraser River, probably because their habitat requirements are more restrictive than the more common Chinook salmon and steelhead because they typically require use of a lake. In Idaho, this combination is especially prevalent in the glacial lakes of the Sawtooth Mountains, particularly Redfish Lake where they are monitored, though they historically occurred in other areas (e.g., Payette Lake, where there were reports of commercial harvests of some 77,000 sockeye salmon in the 1870s); however, some lakes still retain the residualized kokanee form. More to the point, drought periods are common in the northern Rocky Mountains and Pacific Northwest and have occurred with some frequency during this 14,000 year period (see e.g. here and here). A similar drought situation to 2015's occurred in 1977 leading to trucking young salmon downstream around some major dams, a process that has been used frequently to help mitigate the threats faced by outmigrating smolts .Without directly comparing historic droughts, we can get an idea of how streamflow fluctuates regularly in the region with this graph of US Geological Survey data from Johnson Creek, a South Fork Salmon River†† tributary in west-central Idaho:

Overall, streamflow has been variable but relatively stable over time with some obvious low flow periods. The drought of 1977 is clearly visible, as is a similar period in 2001 and a sustained low flow period in the 1980s and 1990s.  In 1990, no sockeye salmon were observed at Redfish Lake in the Sawtooth Valley.

So drought is common and sockeye salmon populations have been declining to the point of extinction prior to 1990 when they may have, in fact, become functionally extinct as a natural population in Idaho. How does this relate to climate change? I produced a previous post from a U.S. Historical Climatology Network (USHCN) site on the Little Salmon River at New Meadows, Idaho, where raw temperature data suggested at most minimal warming over 120 years. For this post, I chose to look at temperature records from the USHCN site at Lewiston, Idaho, because it's near Lower Granite Dam. The raw data produce this plot:

This doesn't suggest warming, but in the interests of full disclosure, the SHAP adjusted data reverse the apparent stable to downward trend and show a possible upward trend. Data collected prior to 1947 were decreased, on average, by 2.08°F; from 1947 on, adjustments averaged 0.07°F, mostly upward. Here is the plot:

The reader can choose whether the data adjustments were appropriate and whether this indicates warming over the region, particularly in recent years. This graph shows adjusted data trends since 1979 (a commonly used start date because it starts the "satellite era") and since the "super El Nińo" year of 1998: Maybe slightly up since 1979, maybe down since 1998 but less variable.

In my opinion, climate and drought are stressors that would probably be irrelevant without the other threats to both adult fish and outmigrating juveniles. Adults coping with passage restrictions has already been discussed. Another major impact is water withdrawal, particularly for irrigation: several million acres of cropland are irrigated by impounded water. In addition, reducing streamflow and velocity through impoundment and diversion can increase water temperature irrespective of climate change. Native and non-native predators of juvenile salmon (pikeminnow, smallmouth bass, terns, etc.) have taken advantage of the altered watershed conditions. Sport and tribal fishing still impacts migrating adults. And, ironically, despite all the concern, 45 adult sockeye salmon returned to the Sawtooth Valley in 2015. While fewer than recent years, this is well above the 16 that returned between 1991 and 1998 and the zero return in 1990.

Most scientists realize that the earth has been warming since the Little Ice Age, and for resource management purposes, it's reasonable to assume that trend will continue even if CO2 is largely irrelevant. There is a clear legal requirement to attempt to recover ESA-listed species, and the program with sockeye salmon appears to be working to some extent. Irrespective of climate, however, the question is really whether recovering sockeye in the Salmon River is truly worth the effort; that is, whether people are willing to accept the costs. In a world without tradeoffs, I would like to see the species recover; however, reducing the threats over which people have control would be more directly helpful than worrying about the theoretical threat that is represented by AGW. Hydropower is often regarded as a renewable energy source because the turbines spin so long as the rivers flow. But the Columbia River demonstrates how large an ecological footprint renewable, momentum-based energy production can have. This is environmental destruction on a grand scale; a watershed that can never really be restored. There are movements afoot to remove the lower four Snake River dams, which would surely improve anadromous fish passage to the Salmon River; in fact, some studies have shown that wind and solar could (or already do) more than compensate for reduced power generation from removal of the four Snake River Dams below Lewiston. But others disagree and, in any case, removing those dams, and even others, cannot not restore the cumulative ecosystem damage that has occurred in five states and one Canadian province.

To sum up: The Columbia River has been heavily altered. Some dams may be removed to improve fish passage and efforts will certainly continue to increase anadromous fish numbers. But climate change is currently a theoretical threat in a sea of real, demonstrable threats; to recover species, the demonstrable threats should be addressed first.
___________
* Anadromous fishes spend most of their adult lives in the ocean but spawn and spend some juvenile time in rivers after hatching, then migrating downstream to the ocean.
** 68°F is a mandated threshold established by the National Marine Fisheries to ensure sufficiently cool temperatures during salmon migration.
*** A graphic of the major dams can be found here.
† Sockeye life history and population structure is complicated. I have used this paper to establish the Wisconsinian date.
†† Sockeye salmon likely did not regularly use this stream, but it has a complete flow record and should be approximately representative of Snake River tributaries in Idaho that did.