The San Dimas Experimental Forest (SDEF) is located in southern California and is representative of the chaparral shrublands of the Southwest USA. Chaparral - including genera of Ceanothus, Adenostoma, Quercus, Salvia and Arctostaphylos - is a dense, drought-tolerant vegetation assemblage with a closed canopy 3-5 m in height. Chaparral is a fire-prone ecosystem and wildfires have burned the SDEF about every 40 years. The SDEF was established in 1933 to quantify the water cycle in a steep, semiarid landscape. Study catchments range in size from 15 to 1160 ha and measurements of stream runoff are made in a nested weir and flume arrangement to account for the very flashy flows. Apart from native chaparral vegetation, streamflow measurements in these study watersheds have also quantified the hydrologic response of vegetation type-conversion and fire. Innovations in hydrologic monitoring developed on the SDEF include a critical depth flume (the San Dimas flume) and tilted rain gages to better sample precipitation in mountainous terrain. Subsurface runoff and plant water relations have been measured in a large lysimeter complex. Water quality monitoring shows that stream water in the SDEF has very high levels of nitrate, derived from atmospheric deposition of chronic air pollution, that approach the Federal EPA standard of 10.0 mg L-1 for nitrate-N. Spreadsheets of rainfall and streamflow (from 1938 to 2015) - the San Dimas Experimental Forest hydrologic database - may be found at the right-hand side of the web page at . Hard copy charts, tables and other records associated with the foregoing data streams are available from the USDA Forest Service, Pacific Southwest Research Station, 4955 Canyon Crest Drive, Riverside, California, 92501 USA or at pete.wohlgemuth@usda.gov.

Southern California chaparral environments, with steep mountain slopes, semi-arid climate, and non-cohesive soils, are very erosive landscapes. Wildfire is the dominant ecological disturbance event in chaparral and it greatly accelerates flooding and erosion, which are directly and/or indirectly related to the loss of the protective vegetation. Since the 1920s, dams and debris basins have been constructed by public works agencies to protect the growing population and infrastructure of southern California by intercepting and impounding flows of water and debris. Dams also capture stream runoff for supplying water to downstream agriculture and urban populations. Major sediment inflows into dams and debris basins following fire can reduce capacity and threaten the ability to provide flood control and water supply. Chaparral provides physical ecosystem services that aid in flood hazard reduction, sediment retention, and the supply of water as well as protecting habitat for endangered species and soil quality.

This chapter provides an overview and comparison of factors influencing hydrological processes, especially stream flow dynamics and evapotranspiration, at ten relatively undisturbed reference watersheds in the USDA-EFR network. The chapter demonstrates the breadth of the hydrogeological, topographic, climatic and ecological characteristics of reference watersheds by discussing how factors such as climate, topography, geology, soils and vegetation influence runoff generation of these reference watersheds. Further, the chapter also briefly considers how site factors influence evapotranspiration, which determines water balance and regulates stream flow.

Agricultural straw, hydromulch, and wood shred or wood strand mulches increasingly are being used as post-fire hillslope treatments, but the differences in effectiveness among these mulch treatments are not fully understood. Following the 2002 Hayman fire in central Colorado and the 2003 Cedar fire in southern California, matched catchments were monitored for five to seven post-fire years to determine the effectiveness of wheat straw mulch (Hayman fire only) and hydromulch in reducing post-fire runoff, peak flow rates, and sediment yields from natural rainfall. Measured runoff and sediment yields were caused by short duration high intensity summer storms at the Hayman fire and long duration winter rains at the Cedar fire.The wheat straw mulch treatment significantly reduced peak flow rates and sediment yields at the Hayman fire. The annual peak flow rates in the first two post-fire years in the straw mulch catchment were 4.5 and 3.9 m(3) s(-1) km(-2) (respectively) as compared to 4.3 and 7.1 m(3) s(-1) km(-2) (respectively) in the control. In post-fire years one and two, the maximum event sediment yields in the straw mulch catchment were 7.2 and 10 Mg ha(-1), respectively, which were less than half of the maximum event sediment yields in the control catchment (19 and 24 Mg ha(-1), respectively). The straw mulch catchment had no detectable runoff or sediment yield after the second post-fire year, but the control catchment continued to have measurable runoff and sediment yields through the seventh post-fire year. The straw mulch treatment effect in runoff reduction was not Significant in the statistical model. Total ground cover was 80% immediately after the application of straw mulch, and decreased to 10% by the end of first post-fire year, yet total ground cover values remained high as litter and vegetation, including invasive cheatgrass, increased.The hydromulch cover at both fires declined rapidly and provided less than 10% of the ground cover within 2.5 months after application at which point the catchment was presumed to be untreated. Due to differences in precipitation, the three catchments at the Cedar fire had significantly different hydrologic responses during the presumed untreated portion of the study, which precluded evaluation of treatment effectiveness during the short treated period. The peak flow responses from the hydromulch and control catchments at the Hayman fire were also different during the presumed untreated period and were not tested. Although the runoff and sediment yields did not differ during the presumed untreated period and were tested for treatment effects, the Hayman hydromulch treatment did not significantly affect either response during the first post-fire year the presumed treated period.Unit-area sediment yields from the catchments were similar to those measured on hillslope plots at both the Hayman and Cedar fires in the first post-fire years, but in later years the sediment yields from the catchments were at least double the sediment yields measured on hillslope plots. The longer periods of greater erosion rates in the catchments likely reflect the addition of channel erosion processes and a difference in hydrologic connectivity at the catchment scale. Published by Elsevier B.V.

In 2002, the Williams Fire burned >90 % of the San Dimas Experimental Forest, providing an opportunity to investigate differences in soil water repellency, peak discharge, and sediment yield between grass-converted and chaparral watersheds. Post-fire water repellency and moisture content were measured in the winter and summer for four years. Peak discharge was determined using trapezoidal flumes with automated stage-height recorders. Sediment yields were measured by making repeated sag-tape surveys of small debris basins. Other than the high summer 2005 increase in repellency on the grass watersheds, only small differences in repellency were observed between the grass and chaparral sites. In general, soil water repellency increased with depth, decreased with time following the fire, and was inversely related to soil moisture content (i.e., least repellent during the winter and most repellent during the summer). Reduction in repellency occurred at moisture contents ranging between 8 % to 16 %. Approximately 85 % of the sediment delivered to the debris basins occurred during the first year, with first year sediment yields being greatest in the chaparral watersheds. Peak discharge was similar for both the grass and chaparral watersheds and was highest following the record rainfall of the 2005 hydrologic year. However, only minor sedimentation followed the record rain events and was similar in both watershed types, suggesting that percent plant cover was sufficient and that the supply of easily mobilized sediment and ravel was depleted after the first post-fire winter.

Following the Cedar Fire (one of seven large wildfires that burned in southern California during the autumn of 2003), aerial hydromulch was applied at 50 and 100% cover to reduce hillslope erosion in chaparral shrublands. Our objectives were to determine the effectiveness of hydromulch in preventing erosion, and to see if plant recovery was hindered by treatment. We installed 54 silt fences to measure sediment production. Five 1-m2 grids were placed behind each fence to measure plant recovery. Hydromulch was effective in reducing erosion immediately after the fire; however, its benefits appeared to be limited to the first 2–4 months following fire, raising doubts as to its overall cost-effectiveness. The rapid breakdown of the hydromulch during the first 6 months after the wildfire provided little hillslope protection during the above-average October 2004 storm events. During the October events, both rainfall amount and storm intensity played a role in the magnitude of sediment production. Hydromulch did not affect post-fire plant recovery, with plant cover measuring >60% at all sites less than 2 years following the wildfire. Accelerated growth of chamise and forbs was likely due to hydromulch prolonging soil moisture retention. Large accumulations of dead litter following die-off of the herbaceous species could increase dry fuels, thus promoting wildfire and therefore shortening the fire return interval.

In the past decade, wildfires around the world have continued to increase in size, severity, and cost. The number of people living in wildland areas has also increased, putting public safety, homes, roads, public infrastructure, water quality, and valued natural resources at risk from wildfire and secondary fire effects. Major concerns after wildfires are the increased runoff and erosion due to loss of the protective forest floor layer, loss of water storage, and creation of water repellent soil conditions. To reduce the potential postfire erosion and flooding, various postfire mitigation treatments are commonly used on highly erodible areas. We have developed and implemented rapid response approaches to compare treatment effectiveness by monitoring sediment yield and runoff response from hillslopes and small catchments. High-intensity rainfall simulation and concentrated flow (rill) experiments are done to compare treated and untreated areas within a burned area. Small watershed impoundments and/or sediment fence barriers are established within weeks following a forest fire and are monitored for three to five years to measure runoff and sediment yields from natural rainfall. These rapid response protocols allow measurements to be made during the first postfire year when runoff and erosion are likely to be greatest with continued monitoring through the initial recovery period. Our study sites in the Western U.S. encompass a range of rainfall regimes including monsoonal rains in the southwest (Arizona and New Mexico), thunderstorms in the Colorado Front Range and Northern Rockies, and wet frontal systems in Southern California. A paired watershed study that examined the effectiveness of contour-felled log erosion barriers found that runoff, peak flows, and sediment yields were generally lower on the treated sites compared to the control sites for lower intensity storms (storms with less than a 2-yr, 10-min maximum rainfall intensity return interval I10-min); however for higher intensity storms (2-yr or greater return interval I10-min) there was no detectable treatment effect. Mulch treatments (agricultural straw, wood shreds or wood strands) appear to out-perform the barrier-type treatments in reducing erosion and can be effective even for the higher intensity events. Our research results have brought a shift in post-wildfire erosion management strategies. For example, in the 1990’s contour-felled log erosion barriers were commonly used to mitigate hillslope erosion on forested landscapes; but this treatment is seldom used today. Since 2002, mulches, especially agricultural straw, are increasingly used for post-fire hillslope stabilization.

High severity wildfire can make watersheds susceptible to accelerated erosion that may impede resource recovery and threaten life, property, and infrastructure in downstream human communities. Land managers often use mitigation measures on the burned hillside slopes to control post-fire sediment fluxes as the first step in ecosystem restoration and to protect human developments. Aerial hydromulch, a slurry of paper or wood fiber that dries to a permeable crust, is a relatively new erosion control treatment that has not been rigorously field tested in wildland settings. Concerns have been raised over the ability of aerial hydromulch to reduce watershed erosion along with its potential for negative impacts on post-fire ecosystem recovery. Since 2007 we have measured sediment fluxes and vegetation development on plots treated operationally with aerial hydromulch and compared them to untreated controls after three separate wildfires in southern California. These study plots, located on steep slopes with coarse upland soils previously covered with mixed chaparral, were monitored with silt fences to trap eroded sediment and meter-square quadrats to measure ground and vegetation cover. We found that aerial hydromulch did reduce bare ground on the treated plots and that some of this cover persisted through the first post-fire winter rainy season. Aerial hydromulch reduced hillslope erosion from small and medium rainstorms, but not during an extreme high intensity rain event. Hydromulch had no

The short term and long term effectiveness of two different treatment methods, Transcutaneous Electrical Nerve Stimulation(TENS Unit) and Acupuncture, on lower back pain were analyzed. The participant pool consisted of a number of different ethnicities and ages. Twenty three subjects were treated with the TENS Unit and twenty five subjects were treated with Acupuncture. All were treated consistently ten times. Their lumbar flexion, lumbar extension, lumbar left lateral bending, and lumbar right lateral bending were measured before the treatment, right after the treatment, and after the tenth treatment. These measurements were compared to normal measurements and taken as a percentage. The average percent improvement for each group was used to determine the effectiveness of each treatment both in the short term and in the long term. According to the data gathered, there have been significant changes in pain severity (Pƒ0:01) among subjects treated with either method. However, 112 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES 45 et al.: 2010 ABSTRACTS IN PROGRAM ORDER Published by OxyScholar, 2010 the Transcutaneous Electrical Nerve Stimulation appears to have induced a higher average percentage of improvement compared to Acupuncture. Thus, Transcutaneous Electrical Nerve Stimulation has been shown to be more effective in increasing lumbar range of motion and pain severity in comparison to Acupuncture Treatments, in both the short term and long term. 116. REDUCING POLLUTION THROUGH GREEN ROOFS: A PHOTOSYNTHETIC STUDY C.A. Frost. Drennan Biology Lab, Loyola Marymount University, Department of Biology, Los Angeles, CA. This study evaluates indigenous, drought-resistant plants that would optimize the collection and storage of carbon dioxide for green roof applications in Southern California. The purpose is to find the plant that has the fastest photosynthetic rate of the following four native Southern Californian plants: frageria californica, galvezia speciosa, heteromeles argutifolia, and salvia clevelandii. The photosynthetic rate as well as the rate of transpiration was measured by a LI-COR 6400XT Portable Photosynthesis System, in the lab of Dr. Phillipa Drennan at Loyola-Marymount University. Five plants per species were tested to limit the influence of experimental error. Once the temperature and pressure were calibrated, each plant was tested at eight different flux levels: 1,800 micro moles per meter squared per second, then 1,500, 1,200, 900, 600, 400, 200, 100, and zero or no light. The results of the study showed that Frageria californica, the wild strawberry, had the most carbon dioxide processed and therefore the fastest photosynthetic rate: 1.3% more than salvia clevelandii and 40.2% more than galvezia speciosa. Frageria californica also had the steepest slope of the light response curve. Frageria californica’s slope was on average 6.8% greater than galvezia speciosa and 40.1% greater than salvia clevelandii for the illumination range 1 to 100 micro moles per meter squared per second. For the illumination range of 200 to 600 micro moles per meter squared per second, salvia clevelandii had the steepest average slope at 5.4% more than frageria californica. Galvezia speciosa reached its steady consumption of carbon dioxide level at an illumination level of 400 micro moles per meter squared per second, the swiftest of all the plants tested. Frageria californica reached the steady consumption level at 900, and salvia clevelandii at 1,200. 117. HIDDEN IN PLAIN SIGHT: IDENTIFICATION AND DESCRIPTION OF A NEW SPECIES OF SPURILLA WITH A PROPOSED MECHANISM OF SPECIATION Bonnie R. Lei. Walnut High School, 400 North Pierre Road, Walnut, CA 91789; Department of Biological Sciences, California State Polytechnic University, Pomona, 3801 West Temple Avenue, Pomona, CA 91768-4032. Through both morphological and molecular analyses of tropical Atlantic and Pacific populations of Spurilla neapolitana (delle Chiaje, 1941), the first such study conducted in the Aeolidiidae family, a cryptic species was identified in the Bahamas. Maximum likelihood phylogenetic trees (100 bootstrap replicates) constructed from hypervariable 16S rRNA and highly conserved H3 histone gene sequences revealed the Bahamas population of S. neapolitana to be in a separate clade from all other S. neapolitana. Qualitative morphological evaluation of diagnostically reliable features including the external morphology, radulae, and reproductive system through scanning electron micrographs, camera lucida drawings, and digital photography corroborated the genetic findings. Radulae were further analyzed quantitatively with a Mann-Whitney U-Test (P50.01) of cusp and denticle length and width ratios and the statistical data confirmed qualitative observations. With a comparison to S. neapolitana and other aeolid species, this new species is here described with a proposed mechanism for speciation. By utilizing a molecular clock constructed using the Langley-Fitch likelihood method, the divergence of this new species from S. neapolitana was determined to be 5.88 million years ago (mya). Paleogeographic data indicate that during this time, from 5–6 mya, the closure of the Havana-Matanzas Channel restricted ocean current flow in the Caribbean and caused a separation between the Bahamas population and the rest of S. neapolitana, which likely led to subsequent speciation. These results provide crucial insight into the biodiversity of this genus and provide a better understanding of how geological phenomena affect the evolution of marine populations. ABSTRACTS 113 46 Bulletin of the Southern California Academy of Sciences, Vol. 109 [2010], Iss. 2, Art. 7 http://scholar.oxy.edu/scas/vol109/iss2/7 118. AUTONOMOUS ROBOT MODELING OF BACTERIAL MOTION AND CHEMOTAXIS Kenny Lei. Walnut High School, 400 N. Pierre Road, Walnut, CA 91789. Harvey Mudd College, 301 Platt Boulevard, Claremont, CA 91711. Some forms of bacteria have flagella to move through water and have two movement phases: straight swim and random tumbling. Certain forms continuously transition between the movement phases, while other bacteria, such as E. coli, employ chemotaxis to decide between movement phases. Chemotaxis is a process in which bacteria direct themselves toward highly favorable areas of food by either straight swimming when conditions are improving or randomly tumbling to reorient themselves when conditions are unfavorable. A comparison between the completely random movement method and chemotaxis movement was made. In addition, an evaluation of how frequently to check food concentrations and decide between transitioning was completed. Finally, the effects of external disturbances on bacteria were explored. Due to the complexity of creating a bacterial environment with a wide range of food concentrations, a bacterium was modeled using an autonomous robot created with the Arduino platform and programmed in Python. The modeled environment contained a dynamic gradient of light levels that represented food concentrations in a bacterial environment. Findings show chemotaxis movement performs better than completely random movement by an average of 57 percent. Furthermore, checking food concentration levels every 0.5 seconds was the optimal frequency and resulted in the highest average light level. Additionally, external disturbances caused detrimental effects on collecting light for the chemotaxis movement while having minimal effects on the completely random movement. This model demonstrates that random movements are not random: there is a clear evolutionary benefit of chemotaxis movement and the frequent checking of concentration levels. 119. DETECTING ANTIBIOTIC INSENSITIVITIES IN WETLAND BACTERIA Steven Hoffman. Milken Community High School, Los Angeles, CA. Research with: Dr. Gary Kuleck and Dr. John Dorsey, Departments of Biology and Natural Science, Loyola Marymount University, Los Angeles, CA. The purpose of this study was to analyze waterborne bacteria isolated from Del Rey Lagoon, Playa del Rey, Los Angeles for antibiotic insensitivities. Surface water samples were gathered and tested for coliforms and Enterococci levels and later cultured on tryptic soy agar (TSA) non-selective media. These plates were then replicated onto TSA plates containing clinically relevant concentrations of antibiotics. All replica plates were photographed and those plates containing colonies with 2 or more antibiotic insensitivities were recorded. Colonies with 3 or more insensitivities were re-grown on antibioticcontaining plates for validation. This study is a work in progress, with more specific resistance testing and species identification of multiple-resistant colonies in the future. Bacteria harboring multiple antibiotic resistances pose a serious worldwide medical and public health problem. 120. THE EFFECT OF PHOTOSENSITIZATION OF FULLERNE-BASED NANOPARTICLES ON REACTIVE OXYGEN SPECIES GENERATION AND TOXICITY TO VIRUSES IN VARIOUS AQUEOUS ENVIRONMENTS Sumit Mitra. University High School, Irvine, CA 92612, Research with: Center for Environmental Implications of Nanotechnology, Duke University, Durham, NC. With their unique catalytic and degrading capabilities, fullerene based nanoparticles (FNP) have had an increasing presence in consumer and commercial products. However, the potential benefits of these nanoparticles have not been weighed against their environmental implications. When FNPs undergo photoexcitation from the sun they produce toxic reactive oxygen species (ROS). The environmental impact of this toxicity needs to be further understood and therefore this study is the first to: 1) understand the role of FNP aggregation on the production of ROS in synthesized and natural environments 2) understand the role of aging of FNPs on aggregation and ROS production in synthesized and natural environments 3) Assess the toxicity of FNPs on the inactivation of T-7 bacteriophage which are found in aqueous environments.

Following a wildfire in the Santa Ana Mountains of northeast Orange County, California, a monitoring project was established to test whether aerial hydromulch reduced post- fire hillslope and small watershed erosion, and to document its impact on re-growing vegetation. The study site received below normal rainfall both the first and second winters after the fire. A high-intensity thunderstorm at the end of May 2008 produced very high peak rainfall intensities, providing an extreme test for the hydromulch. It appears that the mulch reduced hillslope erosion during low and moderate rainstorms, but not during heavy downpours. Once the hydromulch is removed, the sites are susceptible to erosion during subsequent higher intensity storms. Because the spatial differences in the rainfall were confounded with the treatment and control locations, the effect of the hydromulch on stream channel erosion is unknown. Cover assessments showed that the hydromulch did significantly reduce bare ground on the hillsides, and that this cover persisted until the time of the intense thunderstorm. Differences in pre-fire shrub composition and post-fire herbaceous species composition makes it impossible to separate the hydromulch effects from inherent site differences on plant response, suggesting that there was no treatment effect on vegetation recovery.

See next page for additional authors Follow this and additional works at: http://digitalcommons.unl.edu/jfspresearch Part of the Forest Biology Commons, Forest Management Commons, Natural Resources and Conservation Commons, Natural Resources Management and Policy Commons, Other Environmental Sciences Commons, Other Forestry and Forest Sciences Commons, Sustainability Commons, and the Wood Science and Pulp, Paper Technology Commons

Microbiologist, Hydrologist, Soil Microbiologist, and Geomorphologist, respectively, Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. Department of Agriculture, Glendora, Calif. Abstract: Raindrop impact is a major agent of soil erosion, and any reduction in its effectiveness reduces sediment production from burned watersheds. A chaparral soil treated with heatshock fungi in three combinations was compared to sterile soil to test the ability of postfire heat-shock fungi to diminish raindrop impact erosion. Two to three times as much sediment was detached from sterile soil as from any of the three fungal treatments. A 1-month incubation period was sufficient to allow all fungal treatments to resist raindrop impact.

Fire is a major disturbance event in southwestern ecosystems. A prescribed burn in the San Dimas Experimental Forest provided an opportunity to quantify the effects of fire on soil hydrologic properties and sediment fluxes in chaparral-covered steeplands. Located in the San Gabriel Mountains about 50 km northeast of Los Angeles, a 1.28 ha watershed was instrumented with hillslope sediment collectors and a debris basin. Sediment fluxes were measured for seven years prior to the fire and through the first post-fire winter. Soil samples taken just before and after the fire were analyzed for bulk density, texture, and moisture content. Soil non-wettability was determined pre- and post-fire at the same 105 locations using the water drop penetration test. Post-fire soils were denser, coarser, and much drier than prior to the burn. Although extremely variable, non-wettability generally increased after the fire. Post-fire hillslope erosion was twice as great as pre-burn levels during the dry season and increased by 10-fold in the wet season, despite a record drought year. Post-fire sediment yield was 20 times greater than the unburned annual average. We believe the fire-induced changes in soil properties coupled with reduced ground cover caused an increase in surface runoff that accounts for the large increases in post-fire sediment fluxes.

In southern California, the unrelenting urban expansion into neighboring uplands has created a wildland-urban interface that is increasingly difficult to manage. In September 2002, the Williams Fire burned the San Dimas Experimental Forest (SDEF), mostly at high severity. This event provided an opportunity to describe and analyze the impacts of fire and the historical management practice of type-conversion on post-fire runoff, sediment yield, soil water repellency, and vegetation recovery in chaparral ecosystems at the wildland-urban interface. Results indicate that soil water repellency increased with depth, declined with time since fire, was inversely related to soil moisture, and was only slightly different with the two pre-fire vegetation types. Herbaceous grasses and forbs dominated the post-fire vegetation initially, but all watersheds are reverting back to their pre-fire plant communities. Bare ground declined with time since fire, primarily as the litter layer accumulated. Number of species per watershed was similar with the two pre-fire vegetation types, although the species composition was different. Comparisons revealed similar magnitudes of post-fire watershed response for both pre-fire vegetation types. Runoff was large in the first post-fire year despite only moderate rainfall. Runoff exceeded the measurement equipment during

National Forests in southern California contain fire prone ecosystems and significant watersheds that are susceptible to post-fire erosion. Using the 2,348 ha Kinneloa fire near Pasadena, CA and associated debris basins as a case study, we found that a long time-interval wildfire resulted in $2.5 million of sediment management and watershed rehabilitation costs for several city, county, state, and federal agencies. The wildfire suppression costs were $7 million. There was also $.26 million in lost recreation due to the closure of the area to visitors. The total cost from the Kinneloa fire was $9.721 million. A multiple regression analysis of fire interval and resulting sediment yield indicated that reducing the fire interval from the current average 22 years to a prescribed fire interval of 5 years would reduce sediment yield by 2 million cubic meters in the 86.2 square kilometer watershed adjacent to and including the Angeles National Forest. Cost savings take the form of reduced sediment removal in debris basins reduced need for emergency infrastructure protection structures. Other savings include the lost recreation visitor days due to emergency fire closures of watershed recreation facilities. Direct cost savings to Los Angeles County Public Works in terms of reduced debris basin clean out would be $24 million. The ecological effects of a prescribed fire interval of 5 years were not included in this analysis..

Between 1998 and 2002, six sites were established immediately after large wildfires in the western United States to determine the effectiveness of contour-felled log erosion barriers in mitigating post-wildfire runoff and erosion. In each pair of matched, burned, and small watersheds ( 1-13 ha), one was treated with contour-felled log erosion barriers and one was left untreated as a control. For 4 to 6 post-fire years, runoff and sediment yields were measured and correlated with rain properties. High-intensity rainfall produced most of the measured runoff and sediment yields except in the southern California site, where long-duration rain events produced most of the runoff and erosion. For small rain events ( less than the 2-year return period for the 10-min duration), the runoff, peak flows, and sediment yields were lower in the treated watersheds than in the control watersheds, but there was no treatment effect for rain events with larger return periods. Improper installation and degradation over time reduced the effectiveness of contour-felled log erosion barriers. Rainfall characteristics and installation procedures should be carefully considered before choosing contour-felled log erosion barriers for post-fire hillslope stabilisation.