The Berkeley Science Review: Articles

Swimming upstream
Wine country salmon struggle for survival (view PDF)
by Liza Ray

The classic California meal of salmon and
Chardonnay may also come with a side of irony. The recent rise of vineyards and country homes in Sonoma County may be contributing to the decline in population of the pink-fleshed fish. Adina Merenlender, of the Department of Environmental Science, Policy and Management at UC Berkeley, is trying to predict how continued development near the Russian River in Sonoma County will degrade native salmon habitat. Merenlender's lab recently developed a statistical model that uses patterns of land use and stream quality from the past ten years to predict how habitat degradation will continue. Sonoma is an ideal place for this research because, as Merenlender explains, "there are lots of immediate pressures on the environment there."

According to Merenlender, water management and the number and size of buildings, collectively known as the "development footprint," have the greatest influence on the habitat quality of nearby streams. Development footprint includes structures, roads, and other areas where the ground surface is packed so hard that it is no longer permeable to rainfall. The water aggregates and the runoff becomes like a fire hose, accelerating flow and increasing sediment deposition into nearby streams. "The big issue that most do not appreciate," Merenlender explains of vineyards and other types of development, "is the rerouting of water that causes fast moving water can cause excessive erosion." That excessive erosion eventually accumulates in the spaces between river stones where salmon lay their eggs, depriving the eggs of oxygen. As a result, adult salmon have difficulty producing abundant and healthy offspring.

Before predicting how land use will change salmon habitat quality over the next ten years, researchers first had to quantify how land use is affecting salmon habitat quality in Sonoma today. Merenlender and her team, including the lead author and former postdoctoral fellow Kathleen Lohse, and former graduate student David Newburn, used habitat quality and land use data to establish this relationship. Field crews from the California Department of Fish and Game collected habitat quality data from 922 sites in streams and rivers of Sonoma County. The UC Berkeley scientists gathered land use data available from existing county tax assessor records. From those records and aerial photographs, they categorized land use into three types: urban, rural-residential, and vineyard. The researchers classified a piece of land as "urban" if it had at least one building per acre and as "rural-residential" if a single building occupied between one and forty acres. A piece of land was categorized as "vineyard" if at least ten percent of its area was covered by grape vines.

They discovered that all three land use types are associated with degraded spawning habitat. Per unit area, urban development casts the most ominous shadow. But that's old news––scientists proved long ago that dense urban development leads to poor habitat quality for sensitive fish like salmon. What's novel about this research is that it identifies a strong correlation between poor quality salmon habitat and the abundance of nearby rural-residential developments and vineyards.

The researchers used this strong relationship, in conjunction with the mountain of existing evidence about imposing urban developments, to forecast how changes in land use over the next ten years will affect salmon habitat quality. Unlike suburban sprawl, rural-residential development does not grow adjacent to the urban center. Instead, it's prone to "leap frog" all over a region, and like a leaping frog, it is difficult to predict. The researchers in Merenlender's group built a statistical model that included past growth patterns, regional plans, zoning laws, proximity of lands to major employment centers, availability for development, access to sewer and water service, and, most importantly, quality of salmon habitat associated with a given piece of land. They used all this information and more to predict the likelihood that a given parcel's land use will change salmon habitat quality in the future. Using their model, the researchers predicted that in ten years rural-residential developments and vineyards will cover more than ten times as much land as urban areas.

To boot, the model predicts rural-residential developments and vineyards are more likely to occur in places where good-quality salmon habitat remains. Merenlender says of vineyard and rural-residential land owners, it's "not just the flowpaths on their property, but the cumulative effect also." Those cumulative impacts are precisely what the forecast indicates are the biggest threat to salmon habitat. Merenlender's model is especially useful because it incorporates land value. It identifies parcels that, if conserved, would most cost-effectively preserve salmon habitat. Merenlender explains that habitat conservation groups can help salmon the most by investing in "places that aren't completely degraded but are likely to become degraded."

To curtail the expansion of rural residences and vineyards, Merenlender suggests a shift in supply and demand. She suggests that planners consider the environmental impacts of expanding vineyards and residences, and that people who invest in them consider those impacts as well. Simply put, she says, "You have to get everyone to reduce their footprint. What you want to do is pull everybody in." Creating incentives for high density development, she says, is essential to saving salmon habitat. Perhaps one day soon, people will envision the ideal development as a place where they can walk around their neighborhood, shop, and stop into a restaurant to enjoy their wine and salmon— without the irony.

Liza Ray is a staff member in integrative biology.

Want to know more? Check out:
ecnr.berkeley.edu/facPage/dispFP.php?I=546
and
ispe.arizona.edu/about/people/faculty_details.asp?people_id=684

Comments on this article? Drop us a line at with 'letter to the editor' in the subject!