COMMUNICATIONS Ecological Applications, 14(4), 2004, pp. 969­974 2004 by the Ecological Society of America HOW FINE SEDIMENT IN RIVERBEDS IMPAIRS GROWTH AND SURVIVAL OF JUVENILE SALMONIDS KENWYN B. SUTTLE,1 MARY E. POWER,JONATHAN M. LEVINE,2 AND CAMILLE MCNEELY Department of Integrative Biology, University of California, Berkeley, California 94720-3140 USA Abstract. Although excessive loading of fine sediments into rivers is well known to degrade salmonid spawning habitat, its effects on rearing juveniles have been unclear. We experimentally manipulated fine bed sediment in a northern California river and examined responses of juvenile salmonids and the food webs supporting them. Increasing concen- trations of deposited fine sediment decreased growth and survival of juvenile steelhead trout. These declines were associated with a shift in invertebrates toward burrowing taxa unavailable as prey and with increased steelhead activity and injury at higher levels of fine sediment. The linear relationship between deposited fine sediment and juvenile steelhead growth suggests that there is no threshold below which exacerbation of fine-sediment delivery and storage in gravel bedded rivers will be harmless, but also that any reduction could produce immediate benefits for salmonid restoration. Key words: fine sediment; Oncorhynchus mykiss; Pacific salmonids; parr; river food web; sed- imentation; steelhead trout. INTRODUCTION profoundly affect the emergent ecosystem, particularly Throughout western North America, historically during biologically active periods of seasonal low flow. large populations of native anadromous salmonids are It is during these periods of low flow that demograph- in severe decline or extinct. In the United States alone, ically critical juvenile rearing occurs for salmonids. 26 Evolutionarily Significant Units of Pacific salmonid Despite scientific, political, and commercial moti- are currently threatened or endangered (National Ma- vation to quantify the relationship between fine-sedi- rine Fisheries Service 2003). These declines are in ment loading and juvenile salmon production in river large part attributable to degradation of spawning and systems (and in particular, thresholds beyond which rearing habitat (Nehlsen et al. 1991, Frissell 1993), a impairment occurs), no causal relationship has been major cause of which is increased loading and storage established. Research on the influence of deposited fine of fine sediments (Miller et al. 1989, Bisson et al. 1992, sediment on juvenile salmonids has consisted primarily Waters 1995). of laboratory work and correlative field studies com- The storage of fine sediments (particle sizes 2mm paring salmonid assemblages before and after or up- median diameter) in gravel-bedded rivers is normally stream and downstream of a fine-sediment influx, or a transient phenomenon, as sediments enter and leave among rivers with differing bed compositions. This river channels naturally. Without frequent resupply work has suggested that fine-sediment deposition neg- from upstream sources or termination of gravel mo- atively impacts juvenile salmonids and the food webs bilizing flows, fine sediment is carried downstream to supporting them (Crouse et al. 1981, Murphy and Hall lowland reaches or the sea. Yet anthropogenic activities 1981, Reeves et al. 1993), but field experimental sup- have greatly increased the storage of fine sediment in port has been lacking. As a result, mechanisms by rivers throughout the world. Where it comes to rest in which these effects arise are also poorly understood. river reaches, fine sediment can transform the topog- This is due in part to difficulty in isolating the impacts raphy and porosity of the gravel riverbed in ways that of fine sediment from other co-varying physical factors (e.g., flow velocity and turbulence, channel depth, plan Manuscript received 16 June 2003; revised 17 December form morphology) that can also influence salmonid per- 2003; accepted 17 December 2003; final version received 9 Jan- formance. uary 2004. Corresponding Editor: J. S. Baron. Here we provide the results of an experiment de- 1 E-mail: kbsuttle@socrates.berkeley.edu signed to isolate the impact of fine sediment on a ju- 2 Present address: Department of Ecology, Evolution, and Marine Biology, University of California, Santa Barbara, Cal- venile salmonid (Oncorhynchus mykiss) in its natural ifornia 93106-9610 USA. habitat. We manipulated fine bed sediment in replicate 969 970 KENWYN B. SUTTLE ET AL. Ecological Applications Vol. 14, No. 4 FIG. 1. Juvenile steelhead over 100%, 80%, and 0% embedded substrates (left to right). channels in the South Fork Eel River, California two fish in each channel were chosen to differ in length (39 43 45 N, 123 38 40 W) and measured the growth by 5­9 mm. Fish were confined for 46 d, during which and behavior of steelhead parr and the density and com- time we conducted extensive behavioral observations. position of aquatic invertebrate assemblages on which Observers approached the channels, remained motion- they prey. less for 5 min, and then began 10-min continuous ob- servation periods, in which swimming was distin- METHODS guished from holding and sheltering behavior and all In June 2000, we constructed six 2 m long 1m feeding movements and intraspecific interactions were wide channels in the downstream ends of each of four recorded. A minimum of four such observations per river pools distributed over a 3-km reach. The 24 chan- experimental channel were conducted during the time nels consisted of vinyl flooring secured flush to the fish were confined. riverbed with rebar stakes and metal pipe, with walls At experiment's end, all steelhead parr were mea- protruding 15 cm above the water surface. Each chan- sured, weighed, and released. Any fish that died over nel was assigned to one of six fine-sediment loadings the course of the experiment were immediately re- Communications in a complete randomized block design. We filled each placed. Injury was inferred as the cause of death when channel to a depth of 15 cm with pebbles, gravels, and a fish died after developing fin rot. We observed these cobbles collected from the adjacent riverbed and sifted infections on fish wounded during conspecific attacks. through a 6-mm sieve. The size range of these coarse The infection whitened dorsal or caudal fins, spread framework particles was 6­90 mm diameter. Median from the fin into the body tissue, and resulted in death size class by weight was 22­32 mm diameter. The high- within 2 d. Analyses of growth included only steelhead est sediment loading treatment (``100% embedded- that were enclosed in experimental channels for a min- ness'') received enough fine sediment so that only the imum of 25 d. All growth measurements are reported upper surfaces of the coarse framework particles were on a per day basis. visible. At this level, further additions of fine sediment We sampled the invertebrate community in each would not alter the topography or porosity of the bed channel from sediment cores. Just prior to stocking in a biologically meaningful manner. The other five steelhead in the channels and on the day after steelhead treatments at each site received 80, 60, 40, 20, and 0% were removed, three circular cores (14 cm diameter) of that volume (Fig. 1). Fine sediment consisted of were taken from each channel and pooled. The sedi- particles with diameter 2 mm. Median size class by ment was elutriated to dislodge invertebrates, and all weight was 0.60­1.18 mm diameter. organisms were collected and stored in 70% ETOH. After allowing 25 d for invertebrate colonization and All organisms were classified to family and assigned algal growth, we closed the upstream and downstream to one of three broad functional groups (i.e., burrowing, ends of each channel with 6-mm mesh walls that were armored, and vulnerable) based on life history traits permeable to invertebrate drift and smaller prey fishes influencing availability to steelhead fry. This infor- but did not allow passage of juvenile steelhead. We mation was gleaned from published reports (Merritt and seined steelhead parr from the river and measured and Cummins 1996, Resh et al. 1997) and from direct field weighed each. We then stocked each channel with two observation. Burrowing taxa included oligochaete parr (1 parr/m2), approximating densities in the adja- worms, freshwater clams (Sphaeriidae), one genus of cent open habitat. At stocking, standard length ranged silt-encased Trichoptera (Sericostomatidae: Gumaga), from 37 to 54 mm and mass from 0.65 to 2.53 g. The one family of Megaloptera (Sialidae), two families of August 2004 FINE BED SEDIMENT AND JUVENILE SALMONIDS 971 In addition to reducing prey availability, deposited fine sediments increased steelhead activity. At higher levels of embeddedness, fine sediments filled spaces under and between coarse cobbles, producing a flat and featureless bed. As interstitial refuges and prey de- clined, steelhead spent less time sheltering behind or under cobbles and more time actively swimming (Fig. 4a). Steelhead also exhibited higher levels of intraspe- cific aggression, including attacks (Fig. 4b), as prey availability and visual separation between fish de- creased with higher fine-sediment levels. This likely explains the increased incidence of at least one mor- tality event in more heavily embedded channels (lo- gistic regression, P 0.05, n 24; Fig. 5). FIG. 2. Growth of juvenile steelhead trout in relation to direct manipulation of substratum embeddedness (R2 0.63, DISCUSSION P 0.0001). Changes in growth in mass were similar (R2 Anadromous salmonids have a complex life history 0.59, P 0.0001). Analyses of relative growth, which ac- counts for differences in initial size, and of instantaneous that exposes them to a wide range of threats across growth rate produced similar linear patterns (R2 0.52, P multiple life stages. As a commercially, culturally, and 0.0001 and R2 0.53, P 0.001, respectively). One exper- ecologically valuable group of animals, considerable imental channel with 40% substratum embeddedness con- research effort has been devoted to quantifying the im- tained no fish that survived the minimum 25 d and is thus excluded from the analysis. pacts of these various threats (e.g., dams, fish farms, overharvesting, hatchery fish, invasive organisms, and river and estuarine pollution, degradation, and habitat Communications loss) on wild salmon populations. It is widely known Diptera (Ceratopogonidae and Tipulidae), and one fam- that salmonid stocks decline when land use increases ily of Odonata (Gomphidae). Armored taxa included fine-sediment delivery to gravel-bedded rivers (Bisson two families of snails (Planorbidae and Physidae), two and Sedell 1984, Reeves et al. 1993, Waters 1995), but families of stone-encased Trichoptera (Helicopsychi- mechanistic understanding of the role of fine sediment dae and Limnephilidae), and wood-encased Limne- philidae over 10 mm in length. Vulnerable prey in- cluded three families of wood-encased Trichoptera un- der 10 mm in length (Lepidostomidae, Brachycentri- dae, and Limnephilidae), one family of free-living Trichoptera (Rhyacophilidae), four families of Ephem- eroptera (Heptageniidae, Baetidae, Trichorythidae, and Leptophlebiidae), two families of Plecoptera (Perlidae and Chloroperlidae), three families of Coleoptera (El- midae, Haliplidae, and Psephenidae), two families of Diptera (Blephariceridae and Chironomidae), and three families of Odonata (Aeshnidae, Lestidae, and Coen- agrionidae). Individual insect dry biomass was deter- mined based on length regressions published in the literature or generated in this study. Biomass of bur- rowing organisms was log transformed to meet as- sumptions of regression analysis. FIG. 3. Biomass of invertebrates from sediment core sam- ples taken at the experiment's end (mean 1 SE). There were RESULTS significant linear relationships between fine sediment and the biomass of individual functional groups of invertebrates. As Steelhead growth decreased steeply and roughly lin- fine sediment increased (greater embeddedness), biomass of early with increasing fine-sediment concentration (Fig. vulnerable prey declined (R2 0.42, P 0.001) and biomass of unavailable burrowing organisms increased (R2 0.23, P 2). This result was consistent with the effects of sed- 0.02). A similar pattern was found in the prestocking sam- imentation on the food supply available to steelhead. ples taken on 30 June; there was a significant and negative With increasing fine sediment, invertebrate assemblag- relationship between fine sediment and vulnerable prey bio- es shifted from available prey organisms (i.e., epi- mass (R2 0.35, P 0.003) and a significant and positive relationship between fine sediment and burrowing organism benthic grazers and predators) to unavailable burrow- biomass (R2 0.37, P 0.002). Fine sediment had no in- ing taxa (Fig. 3), so that steelhead confined to channels fluence on the biomass of armored grazers. Similar taxon- with higher levels of sedimentation experienced lower specific responses to fine sediment have been observed in food availability than those in less embedded channels. other studies (Bjornn et al. 1977, Mebane 2001). 972 KENWYN B. SUTTLE ET AL. Ecological Applications Vol. 14, No. 4 Bjornn 1979, Hackelroad and La Marr 1993), where larger drainage areas and lower gradients increase the likelihood of fine-sediment loading and storage. Steel- head trout may be particularly vulnerable, as they re- main in natal streams up to two years longer than other anadromous salmonids. By confining juvenile steel- head over discreet patches of riverbed with experi- mentally imposed fine-sediment concentrations, we were able to investigate the mechanisms underlying previously observed patterns of salmonid declines in response to fine-sediment loading and storage. The decreases in steelhead growth and survival we observed with increasing fine-sediment deposition were associated with lower prey availability and higher activity, aggression, and risk of injury. Declines in growth rates lower survival of salmonids and other fishes (Werner and Gilliam 1984, Walters and Korman 1999). Larger body size confers higher survival of over-wintering (Quinn and Peterson 1996) and smolt- ing (Ward and Slaney 1988, Yamamoto et al. 1999) juvenile salmonids. Recent demographic models indi- cate that these juveniles may be the best age classes to target for effective conservation measures. Even modest reductions in juvenile mortality (i.e., 6­11%) are predicted to reverse population declines in Snake River chinook salmon (Oncorhynchus tshawytscha), regardless of adult dam passage success and egg sur- vival (Kareiva et al. 2000). Differences in growth and FIG. 4. Behavior of steelhead parr in experimental chan- survival imposed by fine sediment could therefore have nels. Data represent mean values for each experimental chan- important population-level impacts. nel. (a) Fish activity (swimming time) is represented by the The flux of fine sediment into and out of river sys- best-fit line from a second-order polynomial regression (R2 tems, while a natural process, has been greatly exac- Communications 0.45, P 0.004). The difference in activity between steel- head in 100% embeddedness channels and those in 0% em- erbated by humans. Land uses that increase erosion, beddedness channels translates to a 47% higher energy ex- particularly road construction (Burns 1972, Megahan penditure, based on metabolic data for the same size class of and Kidd 1972, Beschta 1978, Reid et al. 1981), in- sockeye salmon (O. nerka) under similar environmental con- crease fine-sediment loading, while flow regulation and ditions (Brett and Glass 1973), energy equivalents of animal oxygen consumption (Elliot and Davison 1975), and assum- diversion diminish transport and removal. The steep ing a standard 12-h period of nightly inactivity. (b) Intraspe- dissected terrain and weak parent material of drainages cific aggression is represented by the best-fit line from a along California's North Coast make rivers of this re- second-order polynomial regression (R2 0.56, P 0.0002). gion particularly vulnerable to land-management prac- in these declines has been restricted to the embryo stage. When deposited in riverbeds, fine sediment can re- duce survival of embryos and emergence of fry from redds (nests in the riverbed) by decreasing dissolved oxygen and water exchange and entrapping emerging fry (Chapman 1988). Survival of embryos may not, however, limit salmonid populations. Even where sed- iment influxes destroy many redds, higher survival rates from redds in suitable substrate in heterogeneous reaches could compensate for these losses (Magee et al. 1996). Fry and parr from successful redds must then contend with changes in rearing habitat imposed by fine sediment. Even fry hatched from redds in unim- pacted tributaries and side channels are susceptible, as FIG. 5. Steelhead mortality in experimental channels, in they ultimately rear in larger channels (Reiser and relation to fine sediment. August 2004 FINE BED SEDIMENT AND JUVENILE SALMONIDS 973 tices that increase erosion. Increased storage of fines tershed management: balancing sustainability and environ- dramatically alters river ecosystems. This state is tran- mental change. Springer-Verlag, New York, New York, sient over geomorphic time scales, as fine-sediment USA. Bisson, P. A., and J. R. Sedell. 1984. Salmonid populations pulses move down through steeper, gravel-bedded por- in streams in clearcut vs. old-growth forests of western tions of drainage networks and eventually are dis- Washington. Pages 121­129 in W. R. Meehan, T. R. Mer- charged into lowland floodplains, estuaries, and the sea. rell, Jr., and T. A. Henley, editors. Fish and wildlife rela- In addition, certain steep microhabitats may not retain tionships in old-growth forests: proceedings of a sympo- fine sediments even in rivers receiving heavy loading. sium. American Institute of Fishery Biologists, Morehead City, North Carolina, USA. In this sense, the river is potentially self-cleaning. If Bjornn, T. C., M. A. Brusven, M. P. Molnau, J. H. Milligan, land uses that increase loading or decrease transport of R. A. Klamt, E. Chaco, and C. Shaye. 1977. Transport of fine sediments continue unabated, however, areas of granitic sediment in streams and its effects on insects and formerly suitable juvenile rearing habitat may be lost fish. Bulletin 17, College of Forestry, Wildlife and Range from the riverbed long enough to cause irreversible Sciences. University of Idaho, Moscow, Idaho, USA. Brett, J. R., and N. R. Glass. 1973. Metabolic rates and crit- population declines in resident salmonids. This concern ical swimming speeds of sockeye salmon (Oncorhynchus is particularly important for juvenile salmonids, whose nerka) in relation to size and temperature. Journal of the territoriality limits their ability to crowd into shrinking Fisheries Research Board of Canada 30:379­387. areas of good habitat. Burns, J. W. 1972. Some effects of logging and associated Many states and countries throughout the world have road construction on northern California streams. Trans- actions of the American Fisheries Society 101:1­17. regulations directed at fine-sediment management, in- Chapman, D. W. 1988. Critical review of variables used to tended in part to protect native and introduced salmon. define effects of fines in redds or large salmonids. Trans- None of these regulations derives from known quan- actions of the American Fisheries Society 117:1­21. titative relationships between the amount of loaded or Crouse, M. R., C. A. Callahan, K. W. Malueg, and S. E. stored fine sediment and the performance of salmonids Dominguez. 1981. Effects of fine sediments on growth of juvenile coho salmon in laboratory streams. Transactions in the receiving river. In particular, it is not known of the American Fisheries Society 110:281­286. Communications whether there might be an acceptable level of increase Elliot, J. M., and W. Davison. 1975. Energy equivalents of in fine-sediment loading that causes no damage to sal- oxygen consumption in animal energetics. Oecologia 19: monids or the ecosystems supporting them. Nearly all 195­201. sediment management regulations, however, make this Frissell, C. A. 1993. Topology of extinction and endanger- ment of native fishes in the Pacific Northwest and Cali- reasonable assumption (U.S. EPA 1999). fornia (U.S.A.). Conservation Biology 7:342­353. Our experiment demonstrates that fine-sediment de- Hackelroad, G. R., and T. J. La Marr. 1993. Trapping of position, even at low concentrations, can decrease juvenile steelhead outmigrants from Calf Creek, a tributary growth and survival of juvenile salmonids. We find no of the North Umpqua River. Technical Bulletin 4, Aqua- threshold below which fine-sediment addition is harm- Talk, Region 6, Fish Habitat Relationship. USDA Forest less. These results suggest that any augmentation of Service, Pacific Northwest Region, Portland, Oregon, USA. Kareiva, P., M. Marvier, and M. McClure. 2000. Recovery fine-sediment deposition in steelhead bearing rivers in and management options for spring/summer chinoook this region will further impair this potentially popu- salmon in the Columbia River Basin. Science 290:977­979. lation-limiting life stage, while land management prac- Magee, J. P., T. E. McMahon, and R. F. Thurow. 1996. Spatial tices that decrease fine-sediment loading or storage in variation in spawning habitat of cutthroat trout in a sedi- channels may benefit salmonid populations. ment-rich stream basin. Transactions of the American Fish- eries Society 125:768­779. ACKNOWLEDGMENTS Mebane, C. A. 2001. Testing bioassessment metrics: mac- We thank M. Sorensen, J. Sapp, D. Mahler, K. Peer, M. roinvertebrate, sculpin, and salmonid responses to stream Murakami, G. Takimoto, and T. Iwata for assistance in the habitat, sediment, and metals. Environmental Monitoring field; P. Steel and the University of California Natural Reserve and Assessment 67:293­322. System for stewardship and preservation of the study site; J. Megahan, W. F., and W. J. Kidd. 1972. Effects of logging Baron, B. Dietrich, T. Dunne, B. Harvey, T. Lisle, J. Sabo, and logging roads on erosion and sediment deposition from and A. Cundiff for comments on the manuscript; and P. Kar- steep terrain. Journal of Forestry 70:136­141. eiva for valuable discussion. This work was supported by Merritt, R. W., and K. W. Cummins, editors. 1996. An in- Center for Water and Wildland Resources Grant UCAL- troduction to the aquatic invertebrates of North America. WRC-W-906 and National Science Foundation Grant DEB- Third edition. Kendall Hall, Dubuque, Iowa, USA. 961517 to M. E. Power and an Environmental Protection Miller, R. R., J. D. Williams, and J. E. Williams. 1989. Ex- Agency STAR Fellowship to K. B. Suttle. tinctions of North American fishes during the past century. Fisheries 14:22­38. LITERATURE CITED Murphy, M. L., and J. D. Hall. 1981. Varied effects of clear- Beschta, R. L. 1978. Long-term patterns of sediment pro- cut logging on predators and their habitat in small streams duction following road construction and logging in the of the Cascade Mountains, Oregon. Canadian Journal of Oregon coast range. Water Resources Research 14:1011­ Fisheries and Aquatic Sciences 38:137­145. 1016. National Marine Fisheries Service. 2003. Endangered Spe- Bisson, P. A., T. P. Quinn, G. H. Reeves, and S. V. Gregory. cies Act status of West Coast salmon and steelhead. URL: 1992. Best management practices, cumulative effects, and www.nwr.noaa.gov/1salmon/salmesa/pubs/1pgr.pdf long-term trends in fish abundance in Pacific Northwest Nehlsen, W., J. E. Williams, and J. A. Lichatowich. 1991. river systems. Pages 189­232 in R. Naiman, editor. Wa- Pacific salmon at the crossroads: stocks at risk from Cal- 974 KENWYN B. SUTTLE ET AL. Ecological Applications Vol. 14, No. 4 ifornia, Oregon, Idaho, and Washington. Fisheries 16:4­ 407­410 in R. W. Holzenthal and O. S. Flint, Jr., editors. 21. Proceedings of the 8th International Symposium on Tri- Quinn, T. P., and N. P. Peterson. 1996. The influence of habitat choptera. Ohio Biological Survey, Columbus, Ohio, USA. complexity and fish size on over-winter survival and growth U.S. EPA (United States Environmental Protection Agency). of individually marked juvenile coho salmon (Oncorhyn- 1999. South Fork Eel River total maximum daily loads for chus kisutch) in Big Beef Creek, Washington. Canadian sediment and temperature. [Online: www.epa.gov/region09/ Journal of Fisheries and Aquatic Sciences 53:1555­1564. water/tmdl/eel/eel.pdf .] Reeves, G. H., F. H. Everest, and J. R. Sedell. 1993. Diversity Walters, C., and J. Korman. 1999. Linking recruitment to of juvenile anadromous salmonid assemblages in coastal trophic factors: revisiting the Beverton-Holt recruitment Oregon basins with different levels of timber harvest. model from a life history and multispecies perspective. Transactions of the American Fisheries Society 122:309­ Reviews in Fish Biology and Fisheries 9:187­202. 317. Ward, B. R., and P. A. Slaney. 1988. Life history and smolt- Reid, L. M., T. Dunne, and C. J. Cederholm. 1981. Appli- to-adult survival of Keogh River steelhead trout (Salmo cation of sediment budget studies to the evaluation of log- gairdneri) and the relationship to smolt size. Canadian ging road impact. Journal of Hydrology (New Zealand) 20: Journal of Fisheries and Aquatic Sciences 45:1110­1122. Waters, T. F. 1995. Sediment in streams: sources, biological 49­62. effects, and control. American Fisheries Society, Bethesda, Reiser, D. W., and T. C. Bjornn. 1979. Habitat requirements Maryland, USA. of salmonids in streams. Pages 83­138 in W. R. Meehan, Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche editor. Influence of forest and rangeland management on and species interactions in size-structured populations. An- salmonid fishes and their habitats. Special Publication 19. nual Review of Ecology and Systematics 15:393­425. American Fisheries Society, Bethesda, Maryland, USA. Yamamoto, S., K. Morita, and A. Goto. 1999. Marine growth Resh, V. H., J. R. Wood, E. A. Bergey, J. W. Feminella, J. and survival of white-spotted char, Salvelinus leucomaenis, K. Jackson, and E. P. McElravy. 1997. Biology of Gumaga in relation to smolt size. Ichthyological Research 46:85­ nigricula (McL.) in a northern California stream. Pages 92. Communications