sign in sign up
<<
Home , Debris flow

"INSIDE ~ Legacy of Knowledge 2 Two Compared 3 Integration 4 Biological Dynamics 5 Roadbuilding & Logging 5

F I N D I N G 5 issue one I february 1998

ccS~~t;het WC0'w~thiA11vt"~"Lewis Thomas

LESSONS FROM A FLOODED LANDSCAPE

n the soggy morning of February 6, dramatic power to such a huge landscape "This is when the physical work of the 1996, Grant, a fluvialgeomorphol- event, a feel, a smell to it," Grant says. "It landscape gets done," says Grant. "More Oogist at the Corvallis Lab, logged was absolutely the high point of my and of every kind, from on to the Internet to check the gauging career to date, the field experience you the boulders and the trees to the finest stations in and around the H.J. Andrews dream about." , entered the main in those Experimental Forest, about 100 miles 24 hours than will in probably the next Swanson vividly describes the kahwoomp southeast. When he saw that the McKenzie 40 or the previous 30 years." And thus sound of giant boulders rumbling along the was going up 1,000 cubic feet per more dramatic change was wrought upon streambed, and the "Rip City!" experience second every hour, he knew some serious the landscape than will occur again until the of watching whole old-growth trees, with chaos was at hand. next "big one." root wads intact, racing down river chan- Grant and Fred Swanson, a Forest Service nels. He recalls standing beside a flooded "One thing the '96 has really empha- geologist and Andrews Forest project main watching the rapid approach, sized," says Swanson, "is the tremendous leader, headed immediately for the moun- trunk first, of an old-growth tree captured value of Forest Service watershed research tain forest. After all, the last major flood by floodwaters. "The tip of the trunk in hanging together through all the decades event, in 1964, had predated both their lodged in the right at my feet. Then of boredom. It's a tricky balance between a research careers. the force of the took the root wad maniacal persistence in collecting baseline and swung it around to lead downstream. data while very little seems to be happen- Both reacted viscerally to the scene of the The current yanked the treetop out of the ing, and yet responding to the current fash- flood, once they were literally standing in it bank, and on it went." ions in science or the latest management- during the next two days. "There is a policy issue." For it is precisely those boring baseline data, decades' worth of them, that let researchers truly "measure" the events of the flood. Their patterns through the decades of boredom provide a context for understanding major flood events.

far(,twpv~~bcr~ ~cULt-CtI, ~ wO'fth. of t'he-rn" ~ let' v~chef'"1r tv~"~e:J~EW~ of~~ Fred Swanson

Gordon Grant ("Torrents of Change") on mountain streams

ATIle /996 f100d crystal1i:ed sciemists' thinking "bolll f100d effects ,,"d dl'llll/llics. 1

EDITOR'S NOTE

The PACIFIC NORTHWEST RE- SEARCH STATION sen'es society by improving the understanding, use, and managemellt of natural resources. This monthly publication presellts science findings for people who make and influence decisions about managing land.

In ourfirst issue, we describe research that reflects our responsive- ness to natural events and our ability to address issues over time. Floods can bring tragedy in the wake of their destruction. Floods also are a natural A. Floods do the physical work of the landscape. process that has shaped our land- 1 scapes. We,hope that our scientific LEGACY OF KNOWLEDGE BUILDS information helps people make wise FROM THE FLOOD OF '96 choices that influence floods associ- ather than dramatically changing exist- The linkingof all these processes creates what ated with forest lands. ing ideas about flood effects and researchers call a disturbance cascade: some R dynamics, the '96 flood seems to have cascades have a snowball effect and get larger The next issue of Science Findings crystallized many of the hypotheses, according as they go, others act more like an unrolling will examine another aspect of cata- to Swanson. Perhaps its most valuable legacy rug, and dissipate their energy and effects has been to encourage new ways of thinking. quite rapidly. For example, the making of a strophic disturbances. We will debris flow starts with saturated that First, the news story in the natural forest is present findings on how anadromous begin to liquefy. Some flows never make it that floods are not just about a lot of water. into the first small channel, hung up by an old- fish are affected by the dynamics of Yes, there's a lot of water, but all the time growth tree, a lack of content, a failure to aquatic ecosystems. that story is developing, uncountable diverse attract a following. The rug is unrolled, the processes are happening in channels energy spent. As ourfirst issue of Science Findings, and on hillslopes. Many of them connected. this represents a beginning. We will On the hillslopes there are , debris But ifthey do gather enough mass, they'll start flows, quantities of snow absorbing water or taking out some streamside shrubs and logs, be making improvements in content melting at various rates, and interactions increasing momentum and power on their and layout as we go along and between the stream and the road system. In way to the main channel. By this time, they're welcome comments. We also would the channels, there is rising water, moving big enough to take on large stands of alder, wood, and sediment input ranging in size from shift boulders the size of Volkswagen bugs, like to expand our initial distribution. silt through gravel to boulders. Everywhere, add their weight to the force of the flow If you have ideas about improving the transfer of potential to kinetic energy. through the channel. The snowball effect. this publication or names to add to FOR FURTHER READING our mailing list, please contact: Baker, V.R.; Kochel, R.C.; Patton, P.C. eds. 1988. Flood geomorphology. New York: John Wiley and Cindy Miner Sons. 503 p. Pacific Northwest Research Station Grant, G.E., and Swanson, fJ. 1995. Morphology and processes ofl'alleyfloors in moulltain stream, P.O. Box 3890 western Cascades. Oregon. In: Costa. J.E.; Miller, AJ.: Potter, K.W.: Wilcock, P.R., eds. Natural Portland, Oregon 97208 l/Iltropogenic influences infllll'ial geomorphology: the wolmanl'olume. Geophysical Monograph No. 89, (503) 808-2135 Washington, D.C.: American Geophysical Union, 83-101. Jones. J.A.: Grant. G.E. 1996. Long-term stormflow responses to c/earcutting and roads in small and large basins, western Cascades. Oregon. Water Resources Research. 32:959-974. USDA Swanson. F.J.: Dymess. C.T. 1975. Impact of clearclltting and road construction on by iiiiii landslides in the wester/! Cascade Range. Oregon. . 1:393-396. United States Forest Service NOTE: Also see Andrews flood web page at: http://www.fsl.orst.edu/lter/navigafr.htm: click on Table Department of Contents: click on Special Reports: flood 1996. of Agnculture unscathed, The effects of the flood were DISTURBANCE CASCADES not uniform between basins, streams, or even adjacent reaches of the same stream. Are flood effects completely random? No. Consider the flood's perspective, Grant suggests: each flood sees a different land- scape. In their moments of chaos, previous Channel- floods changed the landscape, and in the ValleyFloor decades of boredom, land use and vegeta- Disturbance tion also altered the scene, Processes ;;'f For example, if a reach of stream was Road/Stream gouged down to bedrock by a debris flow Network 30 years ago, it provides less resistance to Interactions subsequent debris flows, which therefore have more available energy to transport both wood and sediment downstream. If accumulations of wood remained along the high flow line of a stream, these "wood " acted as buffers for the riparian areas behind them. If a stand of old-growth firs got knocked down last time, the alder that succeeded in its place was less likelyto withstand the force of a debris flow, and more likelyto contribute wood to the gath- ering mass. Thus the effects of one flood leave their footprints for the next flood. Overall, the '96 flood produced a pattem of irregular disturbance, with greatest changes in small channels affected by debris flows, and in reaches of the that were unconstricted by bedrock and there- fore able to accommodate channel migra- tion. Changes in populations of stream organisms reflected the patchiness, with AElIchj700d .'ee.' II dijleremlalld,"'lIl'l'. Where the I 96.J.flood kllocked d01\'11old gr"'l'th. it I'(/\'l'd till' ,ray large changes in some areas and no .fl'r subsl'quellt older slllllds tofall ill IYY6. detectable change in others. With sufficient force, and speeds of 20 to And yet here is one of the apparent para- 30 miles per hour, even remnant stands of doxes of a big flood event: despite all the WRITER'S PROFILE old-growth along the channel cannot always drama and the devastation, the landslides, Sally Duncan is a science communications withstand the impact of such a channelized the debris flows, and the channel alterations, planner and writer specializing in Forest re- mass, The flood is Intractably at work, some parts of the landscape escape almost source issues. She lives in Corvallis, Oregon.

COMPARING RESPONSES TO TWO FLOODS

and management activities do affect how a flood plays out tlons, on the flood plains and midslope," explains Swanson, "and its across the landscape, in a sense because they intensify natural objective was to move logs efficiently, not to consider landscape L instabilities in the system. As the two activities with the great- effects." Furthermore, there had been no big flood events in the est impact on the landscape, logging and road building come under prevIous 20 years to thoroughly "clean" the system, thus leaving a particularly close scrutiny. Both affect the quantity of woody debris in lot of big trees and logging debris to work major change in the small channels, the frequency of mass movements from hillslopes, stream system. and the interaction between streams and roads or bridges. But by I996. there had been 25 years without much logging or How do we accurately track the effects of management? By road construction, and that big flood just 32 years before. comparisons over time, Witness the differences between the '64 Specifically, in 1964 as much as 15 percent of the basin area had and the '96 floods. clearcuts younger than 15 years, and about 80 km of roads the same age. whereas in 1996 there were only 2 percent of the "In 1964. there had been IS years of fairly constant logging. the clearcuts and less than 20 km of roads less than 15 years old. road system had been built into the watel'shed from lowel' eleva- "Younger plantations are thought to have a HYPOTHESIZED DIFFERENCES IN FLOOD RESPONSE, 1964 & 1996 higher susceptibility to sliding because of reduced root strength and possible effects on HILLSLOPES SMALL STREAMS LARGE STREAMS soil-water movement," says Swanson, so slide WOOD-CHARGED numbers from plantations were lower in the Youngroads- CHANNELS '96 flood. In addition, road building methods Older construction ~ had been modified substantially because of 1964 Many Abundant Manybig Maforchange FLOOD lessons from the '64 flood; road related slides /'" slides fineslash--. debrisflow~of valleyfloors were reduced by about half in 1996. Freshdearcuts. ~ ~ : WOOD-POOR There was simply less PEAK FLOW INCREASES t.. CHANNELS and fewer unstable slopes and roads to Older roads - j ". Newer construction "- contribute to massive structural change. Some, Limited Fewer\ ... ~ocallzed 1996 slides mobile debris flows~change of Another player that differs between flood FLOOD woody debris valleyfloors scenes is snow pack. The timescale of snow Old plantadons ,/ pack registers in days and weeks, rather than

R~~~ progressed, three zones of snow effect were where a deeper snow pack first stored then M,d,-~ ~ ~CI;f\t- apparent a lower zone from 400m to 800m released water. and an upper zone above 1;'~ ~of~fycm1l elevation where melting from a thin, I200m where a very deep pack stored much wet snow cover added its volume to high of the precipitation and buffered the inten- thet '6Lf.flcctL-: v~ veUt,ted.t rainfall:a middle zone from 800m to 1200m sity of the in the upper elevations. ~ were--v~ by cWcut h<:UfiH\;1996. Fred Swanson KEY F t N DIN GS The overall pattern of landslides and streamflows was very strongly influ- years and decades like landslides. By affect- . enced by precipitation intensities, and in some areas by snowpack dynamics. ing the timing and height of peak flows, the .n...... amount of water stored in the snow pack .All floods are strongly affected by preexisting conditions and by the can significantlyexacerbate or diminish flood legacies of human actions and natural events. effects. The February 1996 flood came on ...... the heels of high snowfall, as much as .Floods have a wide variety of unpredictable consequences and different I 12 percent of average. As the flood effects even in neighboring stream reaches.

INTEGRATION IS THE NEW WAVE OF SCIENCE

ny flood leaves behind lastingteaching materials. But the landscape, so if we want to understand it. we'd better have a good lessons from this flood have a significant new aspect to grasp of the whole landscape. And that takes a huge effort." A them: integration. Integrated research, in a sense, tries not to care more about one "In studies of the '64 flood, the focus was mostly on a limited set of landscape response than another, tries to "treat" the whole land- questions in smaller watersheds, but we're now looking at larger scape at once. After all, that's what the flood does. basins, and how material is routed through the whole system," Swanson notes that the focus even of an integrated research effort Swanson explains. There is less focus on just counting landslides, can change within the span of studying a single flood. In '96, their counting road failures, and inventorying channel change. Instead, the original focus was, of course, ecosystem and watershed processes, emphasis is on "How do these pieces connect and so what?" And it but it rapidly tumed to public safety and policy after the five south- ain't easy. em Oregon deaths. Ongoing municipal water supply issues, particu- "We still fall into the trap of having some people look just at a hori- larly out of Salem, kept the pot boiling. "We're seeing floods become zontal view of the landscape via roads, and the people right beside an urban interface issue, linkingpeople with wildland hazards like , them taking just the vertical view via landslides." wild animals, and landslides," he says. In fact, as Grant says, "When you try to build an integrated study so Remember, although researchers' must respond to the information that you can understand floods as systems, the connectivity of all needs of policymakers, someone has to be collecting the baseline these events becomes a kind of tyranny. Far from the old days when data. Hence although the '96 flood triggered new studies, it also we dealt separately with roads, channels, slides, water, or wood, we reinforced the value of long-term monitoring of such variables as now "get it" that the flood doesn't care. It moves through a whole climate, stream flow, channel conditions, and biota.

n lAND MANAGEMENT IMPLICATIONS ·Land use practices and policies have changed dramatically since 1964 and appear to be reducing certain kinds offload hazards. ·New policies under the State Forest Practices rules and the Federal Northwest Forest Plan are barely evident yet on the landscape, so we don't know if they will meet their stated objectives. ·Clearcuts are more susceptible to sliding during thefirst two decades after harvest; roads' susceptibility may dimin- ish ol'er time as well, but slide rates from roads remain substantially higher than they are inforested areas.

AScientists are focusing less on simply counting landslides and road failures, instead they are asking, "How do these pieces connect and so what?" BIOLOGICAL DYNAMICS OF FLOODS

eneralities first. "We need to remember these are dynamic speed and agility,and move into the streambed during high winter systems whether we cut trees or build roads or not" says flows. Pacific giant salamanders are also bottom dwellers, move G Swanson. "We've uncovered countless ways in which mostly by crawling, and are not strong swimmers. ecosystems are attuned to floods and their effects." He refers to the "Trout survival was very similar to winters without major floods," as an example. The idea is that land based Swanson says. "But sculpin and salamander were hammered hard, inputs pulsed into a river during a flood may actually improve the declining by about 65 percent. When flood is sufficient to productivity of the river in terms of vegetation, fisheries, and wildlife. cause movement of gravel and boulders along the streambed, What was the biological story in the flood? "Survival of species is a organisms limited to that habitat may be killed by moving ." direct function of whether and where they can hide, such as rela- Grant adds, "The consequences of floods are not easily summarized tively undisturbed side channels and flood plains, and an organism's as good or bad for humans or organisms. Much depends on the life ability to get to the refuges," says Swanson. cycles and strategies of the various critters." Flood responses of cutthroat trout, sculpins, and Pacificgiant salaman- Most management questions, then, need to be framed openly. ders, the most abundant vertebrates in Andrews streams, tell the What might help or harm particular species? How did the wild story. Trout are strong swimmers capable of moving quickly in the system work, and how did the management overlay affect it? stream, even at high velocity, and take cover along edges or in woody debris during high flows. Sculpins move with considerably less

ROADBUILDING AND LOGGING

spects of the road story were already known in sketchy fash- the combination of flood data, baseline data, and integrated ion: ridgetop, midslope, and floor roads have different research. It does appear that younger plantations play a larger role A impacts on the landscape under flood conditions. Roads can than established stands in activity during floods. The ques- fail in a variety of ways. Roads in different landscape positions have tion becomes a complex study of ecological effects playing out different capacities to become a sediment beside risk and hazard management. source or a sediment sink. for m.e+'\I~ co-me.oCM'\.Clttnet'\I "While no models are perfect, results from The '96 flood brought roads under the ~ !fO; b-ut; I ~ CW\I{Vyever flood studies can contribute directly to our Tennyson mapping of high hazard areas aimed at reduc- research spotlight as landscape elements with ("The Brook") on distinct characteristics, rather than individual ing risks to public safety," Grant says. "case studies" of failure. It also provided more Swanson, too, feels that while they cannot numbers to solidify the story. But what really came clear about claim complete accuracy on causes and effects of each slide, exist- roads was how to ask better questions. ing data can offer extremely useful predictions of high hazard sites. Swanson takes some stabs. "How can we reconfigure roads to do "What we have going in management," Grant concludes, "is a grand tomorrow's job, not yesterday's? Now that road maintenance experiment with no possibility of replication. Surely we are required budgets are severely reduced because of less logging, how can we therefore to learn from events like floods that are episodic. Most get roads to take better care of themselves?" importantly, we need to take some Zen moments, learnable moments, before we rush back in to 'restore' things. Our actions Likewise, logging practices are key players in flood times, and the have consequences in time as well as space." details of their effects are coming to be better understood through

EI BULK RATE POSTAGE + FEES PAlD USDA - FS F N D N G 5 PERMIT No. G-40 u.s. Department of Agriculture PacificNorthwest Research Station 333 S.W. FirstAvenue P.O. Box 3890 Portland, Oregon 97208-3890

Official Business Penalty for Private Use, $300

SCIENTIST'S PROFILES

FRED SWANSON, A Research Geologist GORDON GRANT, a Research with the PNW Research Station, has been Hydrologist with the PNW Research studying landslides and other erosion Station. has been studying rivers for processes in western Oregonfor more more than 15 years. Before that. his than 25 years. Swanson is also a leader interest influvial processes was sparked of the National Science Foundation spon- by a decade-long career as a sored Long-Tenn Ecological Research river guide. His research now focuses on program at the H.J. Andrews the structure and beha~'iorof moulltain Experimental Forest. He is also leader of the Cascade Center of streams. and the effects offorest land use, , floods, and other Ecosystem Management. a research-management partnership disturbances on rivers and watersheds in the Pacific Northwest involving Forest Sen'ice Research. the Willamette National Forest. and elsewhere. He is also a Courtesy Associate Professor of and Oregon State University. Geosciences at Oregon State University.

SWANSON and GRANT can be reached at: E-MAlL ADDRESSES are: Pacific Northwest Research Station/USDA Forest Service [email protected] Corvallis Forest Science Laboratory 3200 S.W. Jefferson Way [email protected]. Corvallis, Oregon 97331

The United States Department of Agnculture (USDA) Forest Service is a diverse organization committed to equal opportunity in employment and program delivery. USDA prohibits discnmlnation In Its programs on the basIs of race. color. national ongln. sex. religion. age. disability. political affiliation. and marital or family status. (Not all prohibited bases apply to all programs.) Persons With disabili- ties who require altel11atlve means of communication of program information (Braille. large print audiotape. etc.) should contact the USDA's TARGET Center at (202)720-2600 (voice and TOO). To filea complaint.write the Secretary of Agnculture.USOAWashlngton.DC 20250. or call(800)245-6340 (voice)or (202)720-1127 (TOO). U5DA ISan equal employment opportunity employer.