The Role of Wildfire in Shaping the Structure and Function of California ‘Mediterranean’ Stream-Riparian Ecosystems in Yosemite National Park

Home , Tenaya, Triple divide

The role of wildfire in shaping the structure and function of California ‘Mediterranean’ stream-riparian ecosystems in Yosemite National Park

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Breeanne Kathleen Jackson

Graduate Program in Environment and Natural Resources

The Ohio State University

2015

Dissertation Committee:

S. Mažeika P. Sullivan, Advisor

Amanda D. Rodewald

Desheng Liu

Copyrighted by

Breeanne Kathleen Jackson

2015

Abstract

Although fire severity has been shown to be a key disturbance to stream-riparian ecosystems in temperate zones, the effects of fire severity on stream-riparian structure and function in Mediterranean-type systems remains less well resolved. Mediterranean ecosystems of California are characterized by high interannual variability in precipitation and susceptibility to frequent high-intensity wildfires. From 2011 to 2014, I investigated the influence of wildfire occurring in the last 3-15 years across 70 study reaches on stream-riparian ecosystems in Yosemite National Park (YNP), located in the central

Sierra Nevada, California, USA. At 12 stream reaches paired by fire-severity (one high- severity burned, one low-severity burned), I found no significant differences in riparian plant community structure and composition, stream geomorphology, or benthic macroinvertebrate density or community composition. Tree cover was significantly lower at reaches burned with high-severity fire, however this is expected because removal of the conifer canopy partly determined study-reach selection. Further, I found no difference in density, trophic position, mercury (Hg) body loading, or reliance on aquatically- derived energy (i.e., nutritional subsidies derived from benthic algal pathways) of/by riparian spiders of the family Tetragnathidae, a streamside consumer that can rely heavily on emerging aquatic insect prey. In addition, I observed minimal changes in the above responses at a subset of study locations in the first summer following the extensive and severe Rim Fire of 2013, although shrub cover at one location burned with low-severity ii fire was significantly reduced. Furthermore, fire frequency at the catchment scale was significantly correlated with fluvial geomorphic characteristics (embeddedness, D50, entrenchment, and width-to-depth ratio) and model-selection results indicated that variability in benthic macroinvertebrate density, catchment-scale fire frequency, and precipitation were important drivers of tetragnathid spider density and trophic position.

Along a gradient of drainage area in two rivers located in the same catchment

(also located in YNP), I also measured wildfire as a potential agent of disturbance within a food-chain length (FCL) framework, where I quantified the relative effects of wildfire characteristics (frequency, timing, and magnitude) as well as classic drivers of FCL including hydrologic disturbance, ecosystem size, and productivity on trophic position

(here, a proxy for FLC) as well as reliance on aquatically-derived energy of/by aquatic benthic insect predators and riparian tetragnathid spiders. Ecosystem size (i.e., drainage area and channel width) received strong support as an environmental determinant of both trophic measures, with variability in flood magnitude emerging as an important mechanism linking ecosystem size and invertebrate trophic responses. Fire metrics were highly correlated with drainage area (positive relationship) as there was greater historic fire extent lower in the catchment. Fire did not emerge as a significant driver of trophic responses by invertebrate predators, however initial evidence suggests that non-linear effects of fire may shed further insight into these relationships.

I also estimated reliance on aquatically-derived energy and trophic position of the

American dipper (Cinclus mexicanus) – a species intimately tied to stream systems for energetic and habitat requirements – in 27 mountain streams of the western slope of the

iii central Sierra Nevada affected by frequent, recent, or severe wildfire. Aquatic birds are considered landscape integrators and are constrained by different ecological processes than aquatic organisms, therefore assessment of the trophic dynamics of aquatic-obligate birds may illuminate divergent patterns and processes related to both fire and food-web dynamics. Model-selection results indicated that dippers occupying territories that were longer and within larger catchments relied more heavily on aquatically-derived energy.

Dipper reliance on aquatically-derived energy was also greater in territories draining catchments with a greater proportion of recent, frequent, or severe wildfire, especially in smaller headwater streams (i.e., < 3rd order) possibily indicating a shift toward greater benthic primary poductivity resulting from removal of the riparian canopy by fire.

Precipitation was a strong predictor of dipper reliance on aquatically-derived energy

(negative relationship), especially in network streams (i.e., > 3rd order) which may be driven by shifts in water quality (e.g., turbidity) or assemblages of available prey items.

For dipper trophic position, these same independent variables received support, but were weaker predictors.

Taken together, these results, combined with the long period of time since fire at some study reaches, indicate support for interactions between wildfire and climate across complex spatial and temporal scales as drivers of both structural and functional responses of stream-riparian ecosystems to fire. Generally, stream-riparian organisms are highly adapted to natural disturbance processes, and there is a growing body of literature that suggests that the occurrence of dynamic, mixed-severity fire regimes may be necessary to maintain ecological function and native biodiversity. However, in this study, precipitation

iv and flood magnitude generally were more influential drivers of riparian spider density and reliance on aquatically-derived energy and trophic position of dippers and tetragnathid spiders, indicating that climate variability and hydrology could outweigh the influence of fire in stream-riparian ecosystems of California’s Mediterranean-type climate.

Dedicated to cowboys, mermaids, stowaways, and monkeys: you know who you are!

Acknowledgements

I would like to thank my advisor Mažeika Sullivan, who has been my mentor for a decade. I would also like to thank my committee members, Amanda Rodewald and

Desheng Liu for their contributions, and Eric Toman for helping me with the proposal and candidacy portion of my program. Thank you to the faculty, staff, and students of

The Ohio State University and the School of Environment and Natural Resources. In

Yosemite National Park I received invaluable support from Gus Smith, Sarah Stock, and

Kent Van Wagtendonk. I would surely never re-emerge from the Yosemite Wilderness without the humor and stamina of Dulcinea Groff and Madeleine Ledford, and I deeply appreciate additional field and laboratory assistance received from Lars Meyer, Katherine

Hossler, Melissa Hickson, Danielle Vent, Kai Zhao, Adam Kautza, and Paradzayi

Tagwireyi. I give special thanks to teachers who inspired me to be a scientist: Jim

Durando, Ron Olowin, Steve Bachofer, Carla Bossard, Steve Takata, Penny Morgan, and

Jeff Braatne and thanks to Adam Sowards for teaching me that the humanities contribute essential knowledge and perspectives that inform our world view where science cannot.

Thanks to my mom, Kathy Jackson, for always saying I am the best daughter ever even when I am a brat; thanks to my husband, Paul Koubek, for inspiring me to live a life of adventure without compromise or apology; thanks to my friends for always making me

vii laugh; and special thanks to my dad, Curtis Jackson, for raising me in a fire truck and being my hero.

viii

Vita

June 2000…………………………………………………………..….Sierra High School

May 2004…………………….B.S. Environmental Science: Earth Science Concentration,

Saint Mary’s College of California

December 2008………………….M.S. Leadership in Physical Education and Recreation,

University of Idaho

December 2009……………………………………………...M.S. Environmental Science,

University of Idaho

Publications Jackson, B.K., S.M.P. Sullivan, C. Baxter, and R. Malison (In press) Stream-riparian ecosystems: mixed- and high-severity fire in DeLaSalla, D. and C. Hanson, editors. The Ecological Importance of Mixed-Severity Fires: Nature’s Phoenix. Elsevier Jackson, B.K. and S.M.P. Sullivan (In press) Responses of riparian tetragnathid spiders to wildfire in forested ecosystems of the California Mediterranean climate region, USA. Freshwater Science.

Jackson B.K., S.M.P. Sullivan, and R. Malison (2012) Wildfire severity mediates fluxes of plant material and terrestrial invertebrate to mountain streams. Forest Ecology and Management. 278, 27-34.

Jackson, B.K. and S.M.P. Sullivan (2009) Influence of fire severity on riparian vegetation heterogeneity in an Idaho, U.S.A. wilderness. Forest Ecology and Management. 259, 24-32.

Field of Study

Environment and Natural Resources

Major field

Ecosystem Science

Table of Contents

Abstract ...... ii

Acknowledgements ...... vii

Vita ...... ix

Field of Study ...... ix

Table of Contents ...... x

List of Tables ...... xvi

List of Figures ...... xxv

Introduction ...... 1

Fire in linked stream-riparian ecosystems ...... 1

Yosemite National Park ...... 3

Dissertation objectives and experimental design ...... 5

Literature Cited ...... 8

Chapter 1: High severity wildfire: benefits for biodiversity and conservation of stream- riparian ecosystems ...... 11

Defining wildfire severity and stream-riparian biotic responses ...... 12

Importance of stream-riparian ecosystems ...... 15

Stream-riparian areas and wildfire severity ...... 20

Time since fire matters ...... 21

Spatial scale matters ...... 22

Responses to a gradient of wildfire severity: evidence from the North American West

...... 23

Physical responses ...... 23

Chemical responses ...... 25

Immediate effects on individuals ...... 27

In-stream biotic response – populations and communities ...... 28

Riparian community and ecosystem responses ...... 33

Primary and secondary production ...... 37

Food-web dynamics ...... 38

Biodiversity, conservation, and management ...... 43

Literature Cited ...... 50

Chapter 2: Antecedent and recent wildfire severity in forested ecosystems of the Sierra

Nevada, California, USA do not result in heterogeneous patterns in riparian vegetation and stream geomorphology ...... 64

Introduction ...... 65

Methods ...... 69

Phase 1: Paired design (2011 and 2012) ...... 69

Phase 2: BACIP experiment (2014) ...... 76

Results ...... 78

Phase 1: Paired design (2011 and 2012) ...... 78

Phase 2: BACIP experiment (2014) ...... 82

Discussion ...... 83

Stream geomorphology...... 84

Riparian vegetation ...... 88

Linkages between riparian vegetation and stream geomorphology ...... 91

Conclusions ...... 92

Acknowledgements ...... 95

Literature Cited ...... 96

Chapter 3: Responses of riparian tetragnathid spiders to wildfire in forested ecosystems of the California Mediterranean climate region, USA ...... 125

Introduction ...... 127

Background ...... 128

Approach ...... 132

Methods ...... 134

xii

Phase 1: 2011-2012 ...... 134

Phase 2: BACIP experiment - 2014 ...... 141

Results ...... 142

Phase 1: 2011-2013 ...... 142

Phase 2: BACIP design - 2014 ...... 144

Discussion ...... 144

Acknowledgements ...... 150

Literature Cited ...... 152

Chapter 4: Variability in invertebrate predator trophic position and reliance on aquatically-derived energy linked to ecosystem size, flood magnitude, and wildfire in a

California Mediterranean-climate river system...... 180

Introduction ...... 182

Methods ...... 186

Results ...... 195

Discussion ...... 200

Reliance on aquatically-derived energy ...... 201

Trophic position ...... 203

Wildfire effects ...... 205

Conclusions ...... 208

xiii

Acknowledgements ...... 209

Literature Cited ...... 210

Chapter 5: Influence of precipitation and wildfire on trophic position and energy sources of American dippers (Cinclus mexicanus) in headwater and network streams of the central Sierra Nevada, California, USA...... 229

Introduction ...... 231

Methods ...... 237

Results ...... 246

Discussion ...... 248

Reliance on aquatically-derived energy ...... 249

Trophic position ...... 251

Conclusions ...... 256

Acknowledgements ...... 257

Literature Cited ...... 258

Appendum B: Paired comparison of tree swallow blood and feces for stable isotope

analysis ...... 280

Methods ...... 280

Results ...... 281

Literature Cited ...... 285

xiv

Conclusion ...... 286

Bibliography ...... 289

APPENDIX A: Geographic Coordinates of all Study Locations ...... 317

APPENDIX B: Stream Geomorphology and Sediment Data ...... 322

APPENDIX C: Riparian Vegetation Data ...... 335

APPENDIX D: Benthic Macroinvertebrate Data ...... 360

APPENDIX E: Chemical Water-Quality Data ...... 363

APPENDIX F: Stable Isotope and Contaminant Data ...... 371

APPENDIX G: Rapid Habitat Assessment (RHA) Data ...... 383

APPENDIX H: Independent Variables Used in Model Selection ...... 386

APPENDIX I: Conceptual Diagram of Environmental Drivers of Food-Chain Length 389

APPENDIX J: Supplementary Maps ...... 391

APPENDIX K: National Park Permit ...... 398

List of Tables

Table 1. Stream reaches in Yosemite National Park, California were paired by fire severity with an attempt to minimize variation in time since fire, elevation, aspect, stream channel morphology, and dominant vegetation. Last year burned refers to the last time the study reach burned; in some cases there was a more recent fire elsewhere in the catchment. Channel-reach morphology was classified following Montgomery and

Buffington (1997). Stream order was determined based on Strahler (1957)...... 103

Table 2. Correlations (Pearson’s r) of physical parameters (elevation, stream gradient, and stream geomorphic measures), time since fire, and individual plant species importance with NMS axes representing a 3-dimensional representation of relative riparian woody and riparian herbaceous species importance in community space. Bold- faced values represent r > 0.5. Only species correlations greater than 0.5 are shown. .. 105

Table 3. Results of paired before-after control-impact (BACIP) analysis of riparian herb, shrub, and tree cover, taxa richness, D50, embeddedness, width-to-depth ratio, entrenchment, and incision at two stream reaches burned by the Rim Fire and two control reaches. Change in means and t-test results are presented. See text for study reach descriptions...... 107

Table 4. Appendum A: Common herbaceous and woody species and their occurrence at each study reach. Site codes are as follows: BV - Buena Vista, FR - Frog, MT - Middle

Tuolumne, TA - Tamarack, CR - Crane, ME - Meadow, MO - Mono, CS - Cascade, ST - xvi

South Tuolumne, GR - Grouse, CY - Coyote, CA - Camp, CH - Chilnualna. Chilnualna

Creek is classified as a reference stream. It was sampled in 2014 following the Rim Fire, but was unaffected by the fire, and has not burned significantly anywhere in the catchment for > 80 years...... 109

Table 5. Stream reaches from Yosemite National Park were paired by fire severity with an attempt to minimize variation in time since fire, elevation, aspect, stream channel morphology, and dominant vegetation. “Year burned” refers to the last time the study reach burned; in some cases there was a more recent fire elsewhere in the catchment.

Channel-reach morphology was classified following Montgomery and Buffington (1997).

Stream order was determined based on Strahler (1957)...... 159

Table 6. Data for each study reach (by low-severity and high-severity fire pair) for tetragnathid spider responses, benthic macroinvertebrates, shoreline habitat, stream geomorphology, precipitation, catchment size, and fire frequency (proportion of catchment burned > 2x since 1930) and fire extent (proportion of catchment burned with moderate-to-high severity fire) at the 12 paired reaches surveyed in 2011 and 2012. TP is trophic position. SD is standard deviation. Hg is mercury concentration. EPT indicates percent of the benthic macroinvertebrate community from the orders Ephemeroptera,

Plecoptera, and Trichoptera. D50 is median sediment size. Precipitation (sampling year indicated in parentheses) was calculated as the average of each monthly total between 01

Oct of the previous year to 30 Sept of the year sampling occurred...... 163

xvii

Table 7. Retained regression models (ΔAICc ≤ 4) with corresponding AICc scores,

2 Akaike weights (wi), and variation explained (R ). Null models (i.e., intercept only) are also included. See text for description of independent variables...... 169

Table 8. Results of paired before-after control-impact (BACIP) analysis of tetragnathid spider density, benthic macroinvertebrate density, median sediment size (D50), and shoreline habitat (% wood – small and large, % overhanging vegetation) at two stream reaches burned by the Rim Fire and two control reaches. Change in means and t-test results are presented. See text for study reach descriptions...... 171

Table 9. Fire axis generated from principle components analysis of catchment wildfire variables: eigenvalue and the percent variance captured by the generated axis, along with principal component loadings and the proportion of the variance (r2) each variable shared with the PCA axes...... 217

Table 10. Minimum, maximum, mean ( ), and standard deviation (SD) of predictor variables relating to ecosystem size (drainage area and channel width), productivity

(NDVI), and disturbance (flood magnitude and fire frequency, timing, and severity) for study reaches along a gradient of drainage area size extending from the headwaters of the

Merced River and South Fork of the Merced River to the 4th-order mainstem of the South

Fork of the Merced River and 5th-order mainstem of the Merced River...... 218

Table 11. Mimimum, maximum, mean ( ), and standard deviation (SD) of response variables including reliance on aquatically-derived energy (i.e., proportion of nutritional subsidies derived from benthic algal pathways) and trophic position of tetragnathid spiders and benthic macroinvertebrate predators along a gradient of drainage area size

xviii extending from the headwaters of the Merced River and South Fork of the Merced River to the 4th-order mainstem of the South Fork of the Merced River and 5th-order mainstem of the Merced River. In addition, z and p values from spatial autocorrelation analysis using Moran’s I are presented for each response variable...... 219

Table 12. Retained regression models (ΔAICc ≤ 4) with corresponding AICc scores,

2 Akaike weights (wi), and variation explained (R ). Null models (i.e., intercept only) are also included. Flood is the magnitude of a 50 AEP flood (a flood that has a 50% chance of occurring each year). Drainage area is in km2 and channel width is in m. Fire axis is the first principal component describing fire severity, fire frequency, and fire timing.

NDVI is an index of vegetation greenness derived from 1-m2 resolution National

Agriculture Imagery Program (NAIP) remote imagery...... 220

Table 13. Fire axis generated from a principle component analysis of catchment wildfire variables: eigenvalue and the percent variance captured by the generated axis, along with principal component loadings and the proportion of the variance (r2) each variable shared with the PCA axes...... 266

Table 14. Description of dipper breeding territories in Yosemite National Park,

California, USA sampled in 2012 and 2013. Catchments are Merced River, Tuolumne

River, and South Fork of the Merced River. Drainage position indicates whether a dipper territory is on a tributary, the mainstem, or at a confluence of a tributary and the Merced,

Tuolumne, or South Fork of the Merced Rivers. Stream order is based on Strahler (1952).

...... 267

xix

Table 15. Descriptive statistics of independent and dependent variables used to assess the influence of wildfire, precipitation, and ecosystem size (drainage area and territory length) on dipper reliance on aquatically-derived energy and trophic position. Fire frequency is the proportion of each catchment draining to a dipper territory that has been burned 1, 2, 3, 4, or > 4 times since 1930. Fire severity is the proportion of each catchment burned with low (dNBR = 1-2), moderate (dNBR = 3), or high (dNBR ≥ 4) severity since 1984. Severity of overlapping fires were summed for each catchment. Fire timing is the proportion of each catchment burned in the last 5, 10, 20, or 30 years. N =

27 for all variables except Hg (n = 18)...... 270

Table 16. Retained regression models (ΔAICc ≤ 4) with corresponding AICc scores,

2 Akaike weights (wi), and variation explained (R ). Null models (i.e., intercept only) are also included. Drainage area is in km2 and territory length is in m. Fire is the first axis of principal component analysis describing fire severity, fire frequency, and fire timing.

Precipitation is average monthly precipitation for the water year corresponding to the sampling year...... 272

Table 17. Mean ( ), standard deviation (SD), and results of paired t-test comparisons of nutrient and stable isotope composition of blood and feces samples collected from riverine swallows occupying nest boxes along the Scioto River, Columbus, OH, USA in

2014. Percent C and δ13C values did not differ between blood and feces, however %N was significantly lower and 15N significantly more depleted in feces samples compared to blood samples...... 283

Table 18. Latitude and longitude of (1) 12 paired reaches and reference reach used in

Chapters 2 and 3, (2) 31 study reaches sampled along a gradient of drainage area used in

Chapter 4, and (3) 27 dipper breeding territories sampled in 2012 and 2013 and used in

Chapter 5. All values are expressed as UTM. Map datum: WGS 84...... 318

Table 19. Stream cross-section data for the 12 paired reaches and reference reach on

Chilnualna Creek used in Chapters 2 and 3. Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim

Fire. All distance and elevation measures are in reference to the thalweg (i.e., the thalweg is at 0 m elevation and 0 m distance) and all values are expressed in m...... 323

Table 20. Wolman pebble counts (3 locations per reach) for the 12 paired reaches and reference reach on Chilnualna Creek used in Chapters 2 and 3. All values are expressed as number belonging to each size class. Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim Fire.

...... 327

Table 21. Percent embeddedness for 50 individual sediment clasts at each of 12 paired reaches and reference reach on Chilnualna Creek used in Chapters 2 and 3. Both before and after data are presented for the study reaches utilized in the before-after-control- impact analysis following the Rim Fire...... 334

Table 22. Ocular estimates (octave class) of cover in the herb layer at each of 12 paired reaches and reference reach on Chilnualna Creek used in Chapter 2. Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim Fire. Species codes are the first two letters of the genus

xxi followed by the first two letters of the species (e.g., HEMI is Huechera micrantha; see

Appendum A for a complete list of species). Ocular estimates are based on octave classes: 1 (trace), 2 (0-1%), 3 (1-2%), 4 (2-5%), 5 (10-25%), 6 (25-50%), 7 (25-50%), 8

(50-75%), 9 (75-95%), 10 (>95%)...... 336

Table 23. Ocular estimates of cover in the shrub layer (octave class) at each of 12 paired reaches and reference reach on Chilnualna Creek used in Chapter 2. Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim Fire. Species codes are the first two letters of the genus followed by the first two letters of the species (e.g., BEOX is Betula oxidentalis; see

Appendum A for a complete list of species). Ocular estimates are based on octave classes: 1 (trace), 2 (0-1%), 3 (1-2%), 4 (2-5%), 5 (10-25%), 6 (25-50%), 7 (25-50%), 8

(50-75%), 9 (75-95%), 10 (>95%)...... 348

Table 24. Ocular estimates of cover in the tree layer (octave class) at each of 12 paired reaches and reference reach on Chilnualna Creek used in Chapter 2. Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim Fire. Species codes are the first two letters of the genus followed by the first two letters of the species (e.g., ABCO is Abies concolor; see

Appendum A for a complete list of species). Ocular estimates are based on octave classes: 1 (trace), 2 (0-1%), 3 (1-2%), 4 (2-5%), 5 (10-25%), 6 (25-50%), 7 (25-50%), 8

(50-75%), 9 (75-95%), 10 (>95%)...... 357

Table 25. Density (number of individuals per m2) of benthic macroinvertebrates collected by Surber sampler (500 μm mesh, 2-9 samples per reach depending on channel width) at

xxii each of the 12 paired reaches and reference reach on Chilnualna Creek used in Chapter 3.

Density values are the mean across all samples for each reach. Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim Fire...... 361

Table 26. Chemical water-quality data collected using a YSI Multiparameter Meter,

Cole-Parmer, Vernon Hills, Illinois, USA at each of the 12 paired reaches and reference reach on Chilnualna Creek used in Chapters 2 and 3. All chemical-water quality measures were made in the summer of 2014. NA signifies no data for that sampling date due to logistical constraints during sampling. * No flow signifies disconntinuous surface flow. Note that these data were not used in the analysis, but provide ancillary information...... 364

Table 27. Naturally-abundant stable isotope ratios of carbon and nitrogen of benthic algae, detritus (stream conditioned leaf litter), spiders of the family Tetragnathidae, and predatory benthic macroinvertebrates of the orders Plecoptera and Megaloptera collected at each of the 12 paired reaches used in Chapter 3. “--” indicates that no individuals were found or collected from that study reach...... 372

Table 28. Contaminant body loading of Tetragnathidae expressed as μg kg-1 collected at each of the 12 paired reaches used in Chapter 3. “--” indicates that no individuals were found or collected from that study reach...... 373

Table 29. Naturally-abundant stable isotope ratios of carbon and nitrogen of benthic alage, detritus (stream conditioned leaf litter), spiders of the family Tetragnathidae, benthic macroinvertebrates of the order Plecoptera, as well as other benthic

xxiii macroinvertebrates collected at each of the 31 locations along a gradient of drainage area used in Chapter 4. “--” indicates that no individuals were found or collected from that study reach...... 374

Table 30. Naturally-abundant stable isotope ratios of carbon and nitrogen of benthic algae, detritus (stream conditioned leaf litter), and dipper feces collected at each of the 27 dipper breeding territories identified in 2012 and 2013 and used in Chapter 5. Both raw and corrected (i.e., from the linear relationship between swallow feces and blood samples collected from 11 individuals along the Scioto River in Columbus Ohio) values are shown for dipper feces. See Appendum B and Chapter 5 for complete details on application of correction factor...... 378

Table 31. Contaminant loading of American dipper feces expressed as μg kg-1 collected at each of the dipper breeding territories sampled in 2012 used in Chapter 5...... 381

Table 32. Rapid Habitat Assessment (RHA) scores for American dipper breeding territories broken out by habitat parameter used in Chapter 5. For riverbank and riparian area scores for each bank have been combined into one...... 384

xxiv

List of Figures

Figure 1. Wildfire can be conceptualized along a gradient of fire severity, from unburned to high severity burned. In riparian zones, wildfire severity is typically determined by assessing changes in both the tree canopy as well as understory vegetation. Low-severity wildfires are commonly characterized by intact riparian canopy and patchy and incomplete burning of understory vegetation whereas high-severity wildfires typically burn the canopy and remove most if not all understory vegetation. The photos presented here represent unburned to high-severity burned riparian zones along low-order streams in Yosemite National Park, USA (above) and the River of No Return Wilderness in central Idaho (below) between three and 11 years post-fire and illustrate some of the common responses to wildfire including erosion and inputs of large wood (Clockwise from top-left: Grouse Creek – low-severity burned three years prior; Buena Vista Creek – high-severity burned patch (foreground) 11 years prior with moderate-severity burned in the background; Tamarack Creek – high-severity burned three years prior; Cliff Creek – high-severity burned five years prior; Upper Cabin Creek – low-severity burned five years prior; Burnt Creek – unburned for > 50 years). Spatial patterns of fire-severity can be highly heterogeneous in riparian zones, therefore differences between low, moderate, and high-severity burned areas are often difficult to distinguish especially as time since fire increases. Photos by Breeanne K. Jackson (top row) and Rachel L. Malison (bottom row)...... 14 xxv

Figure 2. Common stream and riparian organisms that may interact with high-severity wildfire - clockwise from top-left: Baetid mayfly, larval form; Baetid mayfly, adult form;

Tetragnathidae spider; Cutthroat trout (Oncorhynchus clarkii); Western red bat (Lasiurus blossevillii); Sierra garter snake (Thamnophis couchii); and at center Lewis’ woodpecker

(Melanerpes lewis). Illustrations by Madeleine Ledford...... 17

Figure 3. Riparian vegetation at Goat Creek, a tributary of Big Creek in central Idaho, ten days after a moderate-severity fire (August, 2006). In the foreground (and throughout the background), stump sprouting of Betula occidentalis can be seen...... 34

Figure 4. Potential effects of high-severity wildfire on aerial insectivorous bats in riparian corridors in the short- to mid-term (1-10 years following fire, although in some cases longer) under open canopy conditions. Solid arrows represent food web pathways, dashed lines represent indirect effects of wildfire via changes in habitat...... 42

Figure 5. Area burned by decade between 1931 and 2012 in Yosemite National Park,

California, USA...... 113

Figure 6. Fire history by decade within Yosemite National Park, California. Paired reaches are Tamarack (1a) and Grouse (1b); Meadow (2a) and Camp (2b); Buena Vista

(3a) and Mono (3b); Middle Tuolumne (4a) and South Tuolumne (4b); Frog (5a) and

Cascade (5b); and Crane (6a) and Coyote (6b). The reference study reach is (r)

Chilnualna...... 114

Figure 7. The 2013 Rim Fire perimeter and paired stream reaches sampled in 2011 and

2012. Middle Tuolumne and South Tuolumne were consumed by the Rim Fire with high- and low-severity wildfire, respectively, and were used as impact study reaches in 2014.

xxvi

Frog and Cascade were unburned by the Rim Fire and used as high- and low-severity control reaches, respectively...... 115

Figure 8. Geomorphic measurements [width-to-depth, entrenchment, and incision ratios,

D50 (mm); and embeddedness (%)] presented by study pair. Light-grey columns represent high-severity burned stream reaches, dark-grey columns represent low-severity burned reaches, and the black column represents the reference reach on Chilnualna Creek. .... 116

Figure 9. Relationships between fire frequency (i.e., proportion of catchment burned more than twice after 1930) with (a) embeddedness (R2 = 0.55, F = 12.3, p = 0.006); (b)

2 2 D50 (R = 0.43, F = 7.67, p = 0.020); (c) entrenchment ratio (R = 0.56, F = 12.50, p =

0.005); and (d) width-to-depth ratio (R2 = 0.40, F = 6.52, p = 0.029)...... 118

Figure 10. Mean (symbols) and standard error (bars) for % cover of herbs, shrubs, and trees between high-severity and low-severity burned study reaches. Note that overlapping species (in space) can result in cover values that exceed 100%...... 120

Figure 11. 2-D representation of NMS ordination of (a) riparian woody vegetation; and

(b) riparian herbaceous vegetation. Reaches classified as high-severity burned are indicated by crosses, low-severity burned by triangles, and the reference reach by a circle.

The three-axis solution shown is a simplification of riparian species importance in community space. Visually there is little partitioning of riparian species importance indicating highly heterogeneous community composition across study locations irrespective of fire-severity classification...... 121

Figure 12. Results of paired before-after control-impact (BACIP) analysis of riparian herb, shrub, and tree cover at two stream reaches burned by the Rim Fire and two control

xxvii reaches. Means are represented by symbols, standard error by bars. The vertical dashed line represents the Rim Fire. * indicates significance at α = 0.05. ** indicates a trend at α

= 0.10. The treatments for each study reach are: Middle Tuolumne – before high, after high; South Tuolumne – before low, after low; Frog – before high, after control; and

Cascade – before low, after control. Note that overlapping species (in space) can result in cover values that exceed 100%...... 122

Figure 13. (a) Stump sprouting of the common upland plant Arctostaphylos manzanita in the side slope adjacent to Middle Tuolumne Creek six months following the Rim Fire.

120 x 180 mm notebook is shown for scale. Photo taken in March, 2014. (b)

Proliferation of riparian vegetation at Middle Tuolumne Creek 13 months following the

Rim Fire. (Photo credit B. Jackson, October, 2014)...... 124

Figure 14. Fire history by decade within Yosemite National Park. Paired reaches are

Tamarack (1a) and Grouse (1b); Meadow (2a) and Camp (2b); Buena Vista (3a) and

Mono (3b); Middle Tuolumne (4a) and South Tuolumne (4b); Frog (5a) and Cascade

(5b); and Crane (6a) and Coyote (6b). Weather stations are White Wolf, Crane Flat,

Wawona, and Mariposa Grove...... 173

Figure 15. The 2013 Rim Fire perimeter and paired stream study reaches sampled in 2011 and 2012. Middle Tuolumne (4b) and South Tuolumne (4a) were consumed by the Rim

Fire with high- and low-severity, respectively and were used as impact study reaches in

2014. Frog (5a) and Cascade (5b) were unburned by the Rim Fire and used as high- and low-severity control reaches, respectively...... 174

xxviii

Figure 16. Average monthly precipitation (cm) for the Crane Flat weather station for each quarter (i.e., October to December, January to March, April to June, and July to

September) of each water year (October to September) in the study period (2010-2014).

...... 175

Figure 17. Results from paired t-tests for tetragnathid spider response variables (a-d), riparian habitat and stream geomorphology (e-h), and benthic macroinvertebrates (i-j) from 12 paired (high-severity/low-severity) study reaches in Yosemite National Park.

Asteriks represent mean and bars represent +/- one standard deviation from the mean. N

= 12 for all tests except for b (n = 11), c (n = 11), and d (n = 10), where insufficient sample was available for stable isotope and Hg analyses...... 176

Figure 18. Relationship between mean trophic position and body loading of mercury (Hg) of tetgragnathid spiders across Yosemite National Park study reaches. The regression slope indicated a trend for all 12 study reaches (gray dots, dashed line: p = 0.100, R2 =

0.30) but became significant with the removal of Cascade Creek (black dot, solid line: p

= 0.0014, R2 = 0.79)...... 179

Figure 19. Locations of 31 study reaches along a gradient of drainage area size from the headwaters of the Merced River (grey catchment) and South Fork of the Merced River

(taupe catchment) in the Clark Range to the boundary of Yosemite National Park (YNP),

California, USA. Fire history for the 30 years prior to sampling is also shown (1983-

2012)...... 222

Figure 20. Reliance on aquatically-derived energy (i.e., proportion of nutritional subsidies derived from benthic algal pathways) by tetragnathid spiders (diamonds) and benthic

xxix macroinvertebrate predators (squares) from the upstream end of the Merced and South

Fork of the Merced Rivers to 4th and 5th order mainstem segments near the boundary of

YNP...... 223

Figure 21. Trophic position of tetragnathid spiders (diamonds) and benthic macroinvertebrate predators (sqaures) from the upstream end of the Merced River and

South Fork of the Merced River to 4th and 5th order mainstem segments near the boundary of YNP...... 224

Figure 22. Constructed path diagrams for each response variable: reliance on aquatically- derived energy by (a) tetragnathid spiders (χ2 = 0.01, p = 0.994, CFI = 1.00, TFI = 1.09,

RMSE = 0.00) and (b) predatory benthic macroinvertebrates (χ2 = 5.23, p = 0.514, CFI =

1.00, TFI = 1.01, RMSE = 0.00); trophic position of (c) tetragnathid spiders (χ2 = 0.02, p

= 1.000, CFI = 1.00, TFI = 1.05, RMSE = 0.00) and (d) predatory benthic macroinvertebrates (χ2 = 7.03, p = 0.426, CFI = 1.00, TFI = 1.00, RMSE = 0.01). Models were based on predicted relationships and further informed by model-selection analysis and ecological plausibility. Each pathway is labeled with a standardized partial regression coefficient indicating the strength of the relationship. One-headed arrows indicate an assumed causal link and two-headed arrows indicate a correlation with no causality implied. The total variation explained by the model is indicated by R2 values...... 225

Figure 23. Trophic position and reliance on aquatically-derived energy (expressed as a proportion) of/by tetragnathid spiders along a gradient of drainage area. Fire and flow magnitude are shown to illustrate non-linear environmental variability that might influence trophic responses. “Fire axis” represents frequency, severity, and timing. More

xxx positive values indicate a greater proportion of the catchment burned by frequent, severe, or recent fire...... 227

Figure 24. Comparison of linear and piecewise linear relationships of independent variables with trophic responses of tetragnathid spiders: (a) Fire axis by reliance on aquatically-derived energy (linear: t = 36.02, p < 0.001, R2 = 0.55, AIC = -71.07; piecewise: t = 40.55, p < 0.001, R2 = 0.82, AIC = -101.56); (b) Drainage area by reliance on aquatically-derived energy (linear: t = 18.23, p < 0.001, R2 = 0.48, AIC = -68.83; piecewise: t = 32.62, p < 0.001, R2 = 0.84, AIC = -108.57); (c) Fire axis by trophic position (linear: t = 45.04, p < 0.001,R2 = 0.42, AIC = 25.73; piecewise: t = 9.50, p <

0.001, R2 = 0.75, AIC = -0.15); and (d) Drainage area by trophic position (linear: t =

26.27, p < 0.001, R2 = 0.43, AIC = 25.46; piecewise: t = 35.56, p < 0.001, R2 = 0.70, AIC

= 5.00)...... 228

Figure 25. Dipper breeding territories sampled in 2012 and 2013 in Yosemite National

Park, California, USA. Fire frequency is included to illustrate the differences in fire history among catchments draining to dipper breeding territories...... 274

Figure 26. The negative relationship between RHA score and dipper territory length was significant (R2 = 0.17, F = 5.41, p = 0.045; y = 3351.0 – 18.51x)...... 275

Figure 27. Linear regression of precipitation (cm month-1) with (a) reliance on aquatically-derived energy; and (b) trophic position by/of dippers: territory length (m) with (c) reliance on aquatically-derived energy; and (d) trophic position by/of dippers: and fire axis with (e) reliance on aquatically-derived energy; and (f) trophic position by/of dippers. Relationships are partitioned by dipper breeding territories located on

xxxi network (i.e., stream order ≥ 3; drainage area > 9 km2 for this study system) and headwater (i.e., stream order ≤ 2; drainage area < 8 km2 for this study system) streams.

The fire axis is a principal component derived from proportion of each catchment influenced by frequent, recent, or severe fire; more positive values indicate greater proportion of the catchment influenced by fire activity. Precipitation is the average monthly precipitation for the water year (October to September) corresponding with the sampling year. Dashed lines represent network streams. Solid lines represent headwater streams...... 276

Figure 28. Linear regression of a) δ15N for blood and δ15N for feces; and b) δ13C for blood and δ13C for feces of riverine swallows collected from nest boxes along the Scioto

River in Columbus, OH in 2014. Dashed lines indicate 95% confidence curves...... 284

Figure 29. A conceptual diagram (used to inform Chapter 4 hypotheses) showing how post-fire food-chain length may be influenced by ecosystem size, disturbance, and resource availability...... 390

Figure 30. Streams (starting at 1st order) and trails in Yosemite National Park, California,

USA...... 392

Figure 31. Fire perimeters between 1984 and 2012 displayed by year burned within

Yosemite National Park, California, USA...... 393

Figure 32. Number of times burned between 1930 and 2012 (e.g., fire frequency) in

Yosemite National Park, California, USA...... 394

Figure 33. Most recent year burned from 1930 to 2012 displayed by decade except for the

2007-2012, Yosemite National Park, California, USA...... 395

xxxii

Figure 34. Fire perimeters from 1984-2012 shown by year within the Merced River and

South Fork of the Merced River catchments, Yosemite National Park, California, USA.

...... 396

Figure 35. Nested drainages along a gradient of drainage area within the Merced River and South Fork of the Merced River catchments, Yosemite National Park, California,

USA...... 397

xxxiii

Introduction

Fire in linked stream-riparian ecosystems

Stream and river ecosystems and their adjacent riparian zones have long been recognized as critical habitat for myriad organisms (Naiman and Decamps 1997). River corridors are dynamic ecotones linking terrestrial and aquatic habitat and are vectors for the transportation of water, nutrients, and organisms across the landscape (Vannote et al.

1980, Junk et al. 1989, Baxter et al. 2005). In addition, a disproportionally high number of threatened and endangered species rely on riparian habitat in the arid and semi-arid

West (Carrier and Czech 1996) where riparian habitat can increase regional species richness (Sabo et al. 2005).

Fire is the dominant disturbance influencing vegetation in arid and semi-arid forests, shrublands, and grasslands of the American West with fire return estimates ranging from 10-25 years (Agee 1993). Riparian and upland forests exhibit similar fire return intervals (Everett et al. 2003, Van de Water and North 2010). Despite the recognition of fire as a key source of disturbance for forested ecosystems of the American

West (Agee 1993), the role of fire in stream-riparian ecosystems is not well understood

(Gresswell 1999, Bisson et al. 2003, Verkaik et al. 2013). The focus of most contemporary studies relating fire to stream and riparian ecosystems has focused on nutrient flows (Spencer et al. 2003, Hall and Lombardozzi 2008), sediment loading

(i.e., Rood et al. 2007), in-stream primary productivity (Minshall et al. 1989, Minshall et al. 1997, Minshall et al. 2003), and benthic invertebrates (Minshall et al. 1989, Minshall et al. 1997, Minshall et al. 2001, Minshall 2003, Minshall et al. 2003, Minshall et al.

2004, Robinson et al. 2005, Hall and Lombardozzi 2008). The role of fire in terrestrial- aquatic food web connectivity is only beginning to receive attention (Spencer et al. 2003,

Malison and Baxter 2010, Jackson et al. 2012). However, because of the effects of wildfire on riparian plant community structure, composition, and distribution (Dwire and

Kauffman 2003, Bêche et al. 2005, Jackson & Sullivan 2009), fire may strongly influence stream-riparian connectivity by altering the dynamic relationships between riparian vegetation and stream ecosystem functions (Sabo and Power 2002, Paetzold et al. 2005), with important implications for both the conservation and management of stream ecosystems in fire-prone areas.

Although fire can also be an important disturbance agent in Mediterranean- climate forests, most studies relating fire to aquatic ecosystems have been conducted in the temperate regions of the western United States (i.e., Robinson et al. 2000, Minshall et al. 2004, Robinson et al. 2005). With the exception of Bêche et al. (2005) and Kobziar and McBride (2006), there has been very little attention paid to the role of wildfire in central Sierra stream ecosystems. The high degree of hydrologic variability characteristic of Mediterranean-type ecosystems (Bonada and Resh 2013) makes it likely that wildfire effects in Mediterranean-type ecosystems will be distinct compared to temperate ecosystems as the influence of wildfire is significantly mediated by subsequent precipitation, storm events, and flow regimes (Arkle et al. 2010, Shakesby 2011). In

2 addition, California wildfire frequency and extent are predicted to increase markedly due to climate change in the coming decades (Westerling et al. 2011). Therefore, the influence of wildfire on aquatic ecosystems in California is of particular interest to managers of Mediterranean-climate forests.

Yosemite National Park

Yosemite National Park, located in the Sierra Nevada of central California, is a

3,027 km2 park of which 95% is designated as wilderness. The regional climate is

Mediterranean-type characterized by annual dry-wet cycles and high interannual variability in precipitation driven to a large extent by the El Nino Southern Oscillation.

The park typically receives 94.5 cm of precipitation annually with 73.7 cm falling as snow. Yosemite supports myriad biota including some threatened and endangered plant and animal species. Endangered species that utilize riparian and aquatic habitat within the park include the Sierra Nevada mountain beaver (Aplondontia rufra), Sierra Nevada yellow-legged frog (Rana sierra), Yosemite toad (Anaxyrus canorus), bald eagle

(Haliaeetus leucocephalus), harlequin duck (Histrionicus histrionicus), western pond turtle (Actinemys marmorata), and notably several species of bats and wetland plants.

Two National Wild and Scenic Rivers (the Tuolumne and Merced Rivers) begin in the park. These rivers flow westward through glacial valleys with high granite walls finally joining the San Joaquin River in the Central Valley. The Merced River drains

4,470 km2 and has an average discharge of 34 m3 s-1 at its mouth, while the Tuolumne basin encompasses 5,076 km2 and discharges 70 m3 s-1 on average. The headwaters of the

Merced River begin in the Cark’s Range at about 2,413 m above sea level and the

Tuolumne begins at the confluence of the Dana and Lyell Forks in Tuolumne Meadows at 2,616 m.

Yosemite National Park has a long history of managing lightning ignitions to restore vegetation structure and composition and protect resources and infrastructure from fires outside the historic range of variability. Federal land managers received new policy guidance in 2009 that allowed use of a “full range of fire management activities”, thereby allowing them to use wildfire as a tool of restoration (NWCG 2009).

Concurrently, funding for fuels treatments in the Department of Interior fire bureaus has become so limited that the only way parks can treat fuels for restoration and protection is through wildfires managed for resource benefit.

On August 17th, 2013 the Rim Fire began outside Yosemite National Park in the

Stanislaus National Forest. Early analysis suggests that the Rim Fire is unprecedented in both scale (3rd largest fire in the history of CA) and intensity (Flores et al. 2013), making it a realization of the mega-fires predicted to increase in the western U.S. (Pyne 2004).

Although the Rim Fire was unprecedented in extent, high-severity patch size, and proportion of high-severity coverage until it reached the Park boundary, once inside the park, and outside of patches of chaparral created by antecedent fires, the fire burned at moderate and low severity with relatively smaller patches of high severity. The long history of fires managed for resources benefit and the recent Rim Fire provide a unique opportunity to examine whether and to what degree antecedent and recent wildfire affect stream-riparian ecosystem function in mountain drainages.

Dissertation objectives and experimental design

Within the following dissertation I present five chapters that address key aspects of wildfire effects on stream-riparian ecosystem structure and function. Four of these chapters (i.e., 2-5) represent original research.

Chapter 1 is a review of the influence of high-severity wildfire on stream-riparian ecosystem responses across levels of ecological organization (from individuals and populations to communities and ecosystems) with recognition that high-severity wildfire results in winners and losers at the individual and population level over relatively short time periods and small spatial scales, but that heterogeneous conditions resulting from wildfire that occur over broader extents of space and time may be beneficial to communities and ecosystems and therefore important for conservation of limited aquatic natural resources. This chapter will be included in a book to be published by Elsevier in

June 2015, The Ecological Importance of Mixed-Severity Fires: Nature’s Phoenix (D.A.

Dellasalla and C.T.Hanson, Eds.).

The objective of Chapter 2 is to examine the influence of low- and high-severity wildfire occurring over the last two decades on two structural elements – fluvial geomorphology and riparian vegetation – of linked stream-riparian ecosystems in

Yosemite National Park, California, USA. To do this I used a paired design (i.e., one study reach classified as high-severity burned, and one classified as low-severity burned;

12 reaches total). In addition, I examined the relative explanatory power of continuous variability in fire frequency and extent at the catchment scale for determining stream

5 geomorphology at the reach scale. Finally, I compared riparian vegetation structure and composition and stream geomorphology before and after the Rim Fire at a subset of study reaches. This chapter has been submitted for publication to the International Journal of

Wildland Fire.

In Chapter 3, I examine the influence of low- and high-severity wildfire on stream-riparian food web characteristics – as measured by density, reliance on aquatically-derived energy, and trophic position of Tetragnathid spiders; a common riparian consumer that can be highly reliant on emerging aquatic insect prey. Again, a paired design was used at the same study reaches examined in Chapter 2. To complement the categorical paired-reach design and to incorporate catchment-scale features, I also assessed the potential influences of a suite of quantitative, continuous variables at both the reach (i.e., geomorphology, riparian spider habitat, and density and community composition of benthic macroinvertebrates) and catchment scales (i.e., fire extent, fire severity, catchment size, and precipitation) on tetragnathid spider responses. In addition, I compared spider responses before and after the Rim Fire at a subset of study reaches.

This chapter has been submitted for publication in a special fire effects issue of

Freshwater Science and at the time of this writing has been recommended for acceptance by the associate editor.

The objective of Chapter 4 is to determine the comparative influence of ecosystem size, productivity, and disturbance (i.e., wildfire and flood magnitude) on the reliance on aquatically-derived energy sources (i.e., derived from benthic algal pathways) and trophic position of predatory benthic macroinvertebrates and riparian tetragnathid

6 spiders along a gradient of drainage area size in two rivers (the Merced River and South

Fork of the Merced). To do this, I used a continuous design. Because the subject of this chapter has substantial implications for both food-web theory and resource management, the target journal for this chapter is Ecology.

In Chapter 5, I explored the influences of wildfire, ecosystem size, and precipitation on trophic dynamics of the American dipper (Cinclus mexicanus), an aquatic-obligate songbird of western North America. Through their foraging and reproductive activities, birds integrate the stream-riparian environment differently than aquatic invertebrates and fish, which have been the focus of work relating fires to stream biota. Relationships between the American dipper and wildfire dynamics may thus illuminate aspects of fire effects that have not been observed in previous studies. Again, I used a continous design to assess the relative explanatory power of wildfire, ecosystem size, and precipitation for determining dipper trophic dynamics. The target journal for this chapter is the Journal of Animal Ecology.

I anticipate that this work will contribute broadly to food-web theory and, more specifically, to the relative influence of wildfire as a terrestrial disturbance in affecting stream-riparian food web dynamics. In addition, I foresee this work furthering current understanding of the effects of wildfire on both structural and functional elements of stream-riparian ecosystems in Mediterranean-type climate forests.

Literature Cited

Agee, J. K. 1993. Fire ecology of the Pacific Northwest forests. Island Press, Washington, D.C. Arkle, R. S., D. S. Pilliod, and K. Strickler. 2010. Fire, flow and dynamic equilibrium in stream macroinvertebrate communities. Freshwater Biology 55:299-314. Baxter, C. V., K. D. Fausch, and W. C. Saunders. 2005. Tangled webs: reciprocal flows of invertebrate prey link stream and riparian zones. Freshwater Biology 50:201- 220. Bêche, L. A., S. L. Stephens, and V. H. Resh. 2005. Effects of prescribed fire on a Sierra Nevada (California, USA) stream and its riparian zone. Forest Ecology and Management 218:37-59. Bisson, P. A., B. E. Rieman, C. Luce, P. F. Hessburg, D. C. Lee, J. L. Kershner, G. H. Reeves, and R. E. Gresswell. 2003. Fire and aquatic ecosystems of the western USA: current knowledge and key questions. Forest Ecology and Management 178:213-229. Bonada, N. and V. H. Resh. 2013. Mediterranean-climate streams and rivers: geographically separate but ecologically comparable freshwater systems. Hydrobiologia 719:1-29. Carrier, W. D. and B. Czech. 1996. Threatened and endangeres wildlife and livestock interactions.in P. R. Krausman, editor. Rangeland Wildlife. Society for Rangeleand Management, Denver, CO. Dwire, K. A. and J. B. Kauffman. 2003. Fire and riparian ecosystems in landscapes of the western USA. Forest Ecology and Management 178:61-74. Everett, R., R. Schelhaas, P. Ohlson, D. Spurbeck, and D. Keenum. 2003. continuity in fire disturbance between riparian and adjacent sideslope Douglas-fir forest. Forest Ecology and Management 175:31-48. Flores, M., C. Kvamme, B. Rust, K. Takenaka, and D. Young. 2013. Rim Fire BAER - Soils Report. USDA Forest Service. Gresswell, R. E. 1999. Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193-221. Hall, S. J. and D. Lombardozzi. 2008. Short-term effects of wildfire on montane stream ecosystems in the Southern Rocky Mountains: one and two years post-burn. Western North American Naturalist 68:453-462. Hicks, J. H., M. S. Wipfli, D. W. Lang, and M. E. Lang. 2005. Marine-derived nitrogen and carbon in freshwater-riparian food webs of the Copper River Delta, southcentral Alaska. Oecologia 144:558-569.

Jackson, B. K. and S. M. P. Sullivan. 2009. Influence of fire severity on riparian vegetation heterogeneity in an Idaho, U.S.A. wilderness. Forest Ecology and Management 259:24-32. Jackson, B. K., S. M. P. Sullivan, and R. L. Malison. 2012. Wildfire severity mediates fluxes of plant material and terrestrial invertebrates to mountain streams. Forest Ecology and Management 278:27-34. Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood pulse concept in river- floodplain systems. Pages 110-127 International large river symposium. Kobziar, L. N. and J. R. McBride. 2006. Wildfire burn patterns and riparian vegetation response along two northern Sierra Nevada streams. Forest Ecology and Management 222:254-265. Malison, R. L. and C. V. Baxter. 2010. The fire pulse: wildfire stimulates flux of aquatic prey to terrestrial habitats driving increases in riparian consumers. Canadian Journal of Fisheries and Aquatic Sciences 67:570-579. Minshall, G. W., K. E. Bowman, B. A. Rugenski, and C. Relyea. 2003. Monitoring of streams in the Payette National Forest 1988-2003: Big Creek and South Fork Salmon tributaries pre- and post-fire. USGS, Payette National Forest. Minshall, G. W., J. T. Brock, and J. D. Varley. 1989. Widlfire and Yellowstone's stream ecosystems. Bioscience 39:707-718. Minshall, G. W., C. T. Robinson, and D. E. Lawrence. 1997. Postfire responses of lotic ecosystems in Yellowstone National Park, USA. Canadian Journal of Fisheries and Aquatic Sciences 54:2509-2525. Minshall, G. W., C. T. Robinson, D. E. Lawrence, D. A. Andrews, and J. T. Brock. 2001. Benthic macroinvertebrates assemblages in five central Idaho (USA) streams over a 10-year period following disturbance by wildfire. International Journal of Wildland Fire 10:201-213. Minshall, G. W., T. V. Royer, and C. T. Robinson, editors. 2004. Stream ecosystem responses following the Yellowstone wildfires: the first 10 years. Yale University Press, New Haven, CT. Naiman, R. J. and H. Decamps. 1997. The Ecology of Interfaces: Riparian Zones. Annual Review of Ecology and Systematics 28:621-650. Paetzold, A. C., C. J. Schubert, and K. Tockner. 2005. Aquatic-terrestrial linkages along a braided river: riparian arthropods feeding on aquatic insects. Ecosystems 8:748- 758. Pyne, S. J. 2004. Tending Fire: Coping with America's Wildland Fires. Washington Island Press. Robinson, C. T., G. W. Minshall, and T. V. Royer. 2000. Inter-annual patterns in macroinvertebrate communities of wilderness streams in Idaho, USA. Hydrobiologia 421:187-198. 9

Robinson, C. T., U. Uehlinger, and G. W. Minshall. 2005. Functional characteristics of wilderness streams twenty years following wildfire. Western North American Naturalist 65:1-10. Rood, S. B., L. A. Goater, J. M. Mahoney, C. M. Pearce, and D. G. Smith. 2007. Floods, fire, and ice: disturbance ecology of riparian cottonwoods. Canadian Journal of Botany 85:1019-1032. Sabo, J. L. and M. E. Power. 2002. River-watershed exchange: effects of riverine subsidies on riparian lizards and their terrestrial prey. Ecology 83:1860-1869. Sabo, J. L., R. A. Sponseller, M. Dixon, K. Gade, T. Harms, J. Heffernan, A. Jani, G. Katz, C. U. Soykan, J. Watts, and J. Welter. 2005. Riparian zones increase regional species richness by harboring different, not more, species. Ecology 86:56-62. Shakesby, R. A. 2011. Post-wildfire soil erosion in the Mediterranean: review and future research directions. Earth-Science Reviews 105:71-100. Spencer, C. N., K. Odeny-Gabel, and F. R. Hauer. 2003. Wildfire effects on stream food webs and nutrient dynamics in Glacier National Park, USA. Forest Ecology and Management 178:141-152. Van de Water, K. and M. North. 2010. Fire history of coniferous riparian forests in the Sierra Nevada. Forest Ecology and Management 260:384-395. Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130-137. Verkaik, I., M. Rieradevall, S. D. Cooper, J. M. Melack, T. L. Dudley, and N. Prat. 2013. Fire as a disturbance in mediterranean climate streams. Hydrobiologia 719:353- 382. Westerling, A. L., B. P. Bryant, H. K. Preisler, T. P. Holmes, H. G. Hidalgo, T. Das, and S. R. Shrestha. 2011. Climate change and growth scenarios for California wildfire. Climatic Change 109:445-463.

Chapter 1: High severity wildfire: benefits for biodiversity and conservation of stream- riparian ecosystems

Breeanne K. Jackson; S. Mažeika P. Sullivan; Colden V. Baxter; and Rachel L. Malison

Abstract: Due to the linked nature of stream-riparian ecosystems, the highly adapted organisms and food webs that inhabit them, and the disproportionate contribution of natural resources from stream-riparian areas, the role of wildfire in these ecosystems may be essential to managing for biodiversity and conservation across landscapes. The diverse array of organisms that inhabit stream-riparian ecosystems respond to wildfire across a gradient of fire severity. In addition, time since fire and annual timing of fires can interact with other sources of disturbance and life-history events to influence stream-riparian systems across levels of ecological organization. The degree of impact wildfire has also depends on spatial aspects including extent and patchiness, as well as interactions with anthropogenic disturbance. For example, if characteristic stream-riparian continuity is disrupted or other natural features are impaired or lost due to human activities, detrimental effects can occur. Generally, stream-riparian organisms are highly adapted to natural disturbance processes, and the occurrence of dynamic, mixed-severity fire regimes may be necessary to maintain ecological integrity and native biodiversity. 11

Defining wildfire severity and stream-riparian biotic responses

Wildfire is an important natural disturbance that has consequences for both structural and functional characteristics of riparian and stream ecosystems (Resh et al.

1988, Gresswell 1999, Verkaik et al. 2013a). More than 20 years of studies now point to a diverse array of responses by stream-riparian organisms and ecosystems to wildfire.

Ecological responses vary along gradients of fire characteristics, including severity, extent, frequency, time since disturbance, and hydrological context (Agee 1993, Arkle et al. 2010, Romme et al. 2011), among others. Although high-severity fire can result in major changes to stream and riparian areas, including erosion and sedimentation, opening of the riparian canopy, inputs of large wood to the stream channel, and changes in water temperature and chemistry, low-severity fire may have little to no effect (Jackson and

Sullivan 2009, Arkle and Pilliod 2010, Malison and Baxter 2010a, Jackson et al. 2012)

(see Figure 1). Stream-riparian biota respond both directly to wildfire as well as indirectly via wildfire-induced changes in physical habitat (Arkle et al. 2010). Land managers often work to keep high-severity fire out of riparian zones using a suite of techniques, including fuels reduction (removal of trees and understory vegetation through mechanical thinning and/or prescribed fire) and suppression (Stone et al. 2010). However, stream and riparian organisms are often highly adapted to disturbances, including floods, drought, and wildfire (Dwire and Kauffman 2003, Naiman et al. 2005), and dynamic fire regimes that operate over time and space may be important in maintaining ecosystem integrity and biodiversity of linked stream-riparian ecosystems (Bisson et al. 2003).

This chapter focuses on the effects of wildfire across a gradient of severity on organisms and processes in linked stream-riparian ecosystems. To address the range of wildfire effects, we concentrate on probable influences of wildfire on both abiotic and biotic characteristics across multiple levels of ecological organization (from individuals and populations to communities and ecosystems). Rather than presenting a complete review of the literature, we describe in relative depth examples of responses associated with each level of ecological organization. We also focus our discussion on the influences of wildfire severity and how these may vary over time, drawing principally on empirical evidence from the North American West, where much science, as well as resource management uncertainty and public dialogue, has been centered on the costs and benefits of wildfire (Pyne 1997, 2004, Hutto 2008). Moreover, we afford particular attention to fire-food web dynamics because food webs are a valuable window into the structure, function, and productivity of linked stream-riparian ecosystems (Wallace et al. 1997,

Power and Dietrich 2002, Baxter et al. 2005) and can provide spatially and temporally integrated perspectives on the effects of wildfire (e.g., Mihuc and Minshall 2005). We conclude with a broad discussion of the potential importance of high-severity wildfire for biodiversity, conservation, and management of stream-riparian ecosystems.

14 from top-left: Grouse Creek – low-severity burned three years prior; Buena Vista Creek – high-severity burned patch (foreground) 11 years prior with moderate-severity burned in the background; Tamarack Creek – high-severity burned three years prior; Cliff Creek – high-severity burned five years prior; Upper Cabin Creek – low-severity burned five years prior; Burnt Creek – unburned for > 50 years). Spatial patterns of fire-severity can be highly heterogeneous in riparian zones, therefore differences between low, moderate, and high-severity burned areas are often difficult to distinguish especially as time since fire increases. Photos by Breeanne K. Jackson (top row) and Rachel L. Malison (bottom row).

Importance of stream-riparian ecosystems

Even though aquatic ecosystems make up only about 2% of terrestrial landscapes, they are disproportionately relied upon by humans for numerous natural resources (Postel and Carpenter 1997). Streams and riparian areas act as conduits, reservoirs, and purification systems for freshwater (Sweeney et al. 2004). Riparian zones sustain unique communities of organisms, contributing >50%, on average, to regional species richness values (Sabo et al. 2005), and a disproportionate number of threatened and endangered species rely on aquatic and riparian habitats (Carrier and Czech 1996), as do many organisms that provide food, medicine, and fiber to humans. In addition, these areas are valued as scenic and utilized for recreation.

The influence of wildfire as an agent of natural selection has resulted in a suite of organisms that exhibit apparent adaptations that make them resistant or resilient to

15 wildfire, and riparian and aquatic organisms are no exception. Because riparian zones are transitional areas (or ecotones) between aquatic and terrestrial habitats, a diverse array of animals are associated with riparian corridors, ranging from aquatic (fish, benthic invertebrates) to amphibious (frogs, salamanders) to terrestrial (riparian birds, mammals, and reptiles), each exhibiting responses to wildfire that vary across gradients of fire severity (See Text Box 1 and Figure 2).

Despite their importance, riparian areas have been degraded worldwide, and in some regions the majority of riparian zones have been lost altogether. For example, in

California’s Central Valley, approximately 99% of historic riparian zones have vanished due to land-use changes (Khorram and Katibah 1984). These impairments are largely due to a legacy of ecosystem degradation, fragmentation, and loss, as well as the expansion of nonnative species. Additionally, although wildfire may be less frequent in riparian versus upland areas, fire disturbance may be more severe in riparian areas given the greater accumulation of fuel that may occur between wildfire events (Everett et al. 2003). Within this context, wildfire is generally viewed with a mix of concern and optimism. On the one hand, there are concerns about the implications of higher water temperatures and increased erosion and sedimentation for conservation of sensitive species and protection of ecosystem services. On the other hand, wildfire can be important in both maintaining biodiversity and ecosystem function (e.g., Arkle and Pilliod 2010) and has been investigated as a potential restoration technique (e.g., Blank et al. 2003).

Figure 2. Common stream and riparian organisms that may interact with high-severity wildfire - clockwise from top-left: Baetid mayfly, larval form; Baetid mayfly, adult form;

Tetragnathidae spider; Cutthroat trout (Oncorhynchus clarkii); Western red bat (Lasiurus blossevillii); Sierra garter snake (Thamnophis couchii); and at center Lewis’ woodpecker

(Melanerpes lewis). Illustrations by Madeleine Ledford.

Text Box 1

Examples of stream-riparian animals that may benefit from high-severity wildfire.

1. Immediate impacts may be negative, but stream invertebrate abundance and biomass are frequently observed to increase in the short- to mid-term following fire

(Minshall 2003, Verkaik et al. 2013a), and production of emerging adult insects (i.e., aquatic insects that emerge from the water as winged adults) can increase as well (Mellon et al. 2008, Malison and Baxter 2010a). Such increases may be accompanied by reductions in species diversity and dominance by insects that are habitat and trophic generalists, drift-dispersers, and that have multi-voltine (having multiple generations per year) life cycles (e.g., Chironomidae, Baetidae) (Mihuc and Minshall 1995, Minshall et al. 2001b). Climate and hydrologic context following wildfire may mediate mid- to longer-term impacts; for instance, Rugenski and Minshall (2014) reported increases in both invertebrate biomass and diversity in wilderness streams of Idaho over five years following severe wildfire during a period of time characterized by reduced peaks in spring floods.

2. Despite a long-standing assumption that high-severity wildfire has negative impacts on stream fishes, in many cases immediate effects on fishes appear slight or recovery of populations occurs rapidly (Rieman et al. 1997, Sestrich et al. 2011), and there is mounting evidence of numerous indirect, positive effects on fish populations that may follow severe wildfire. For instance, the pulse in invertebrate production that can follow severe wildfire (Malison and Baxter 2010a) may provide increased food resources to fish. Even when wildfire is followed by scouring debris flows that may, at least

18 temporarily, cause extirpation of fish from a local stream reach (Howell 2006), the combination of increased downstream transport of sediment and large wood that creates and maintains essential habitat (Bigelow et al. 2007), and increased export of drifting invertebrate prey from such tributaries (Harris et al. In revision-a) may lead to net positive effects on fishes in recipient habitats. The pulse of natural erosion/sedimentation that can occur shortly after high-severity fire can be associated with increases in native fish populations by three or more years post-fire (Sestrich et al. 2011), possibly due in part to enhanced spawning grounds.

3. Streams and their adjacent riparian zones provide important foraging habitat for insectivorous bats (Seidman and Zabel 2001, Russo and Jones 2003, Fukui et al. 2006), where aquatic insects that emerge from the stream as adults can comprise the majority of bat diets (Belwood and Fenton 1976, Swift et al. 1985). The combination of increased emergence of stream insects and removal of the riparian canopy following high-severity fire may provide bats with better foraging conditions (Malison and Baxter 2010b,

Buchalski et al. 2013) (See Text Box 2 for additional details).

4. Many birds that principally occupy riparian areas also rely on trees burned by fire

(i.e., snags) for nesting cavities. For example, in the western USA, Lewis’ woodpeckers

(Melanerpes lewis), a cavity-nester and an aerial insectivore common in riparian zones, have been called “burn specialists” because they tend to be abundant in both recent (2-4 years post fire) and older (10-25 years post fire) high-severity burns (Linder and

Anderson 1998, Vierling and Saab 2004). Lewis’ woodpeckers and other aerial

19 insectivorous birds can also benefit from increases in emergent insects and other aerial insect prey (e.g., Bagne and Purcell 2011) following high-severity fires.

Stream-riparian areas and wildfire severity

Although wildfire occurs across landscape types (i.e., forests, grasslands, deserts), understanding its role in shaping stream-riparian ecosystems is particularly critical given the important ecosystem services they provide. Notably, stream-riparian ecosystems differ from upland environments in moisture regime, topography, microclimate, vegetation, soils, and productivity (reviewed in Pettit and Naiman 2007) and these differences can influence characteristics of wildfire. Fire severity in riparian zones is influenced by a number of factors including aspect, valley entrenchment, structure and composition of riparian vegetation, and stream size (Van de Water and North 2011). The latter is of particular importance because wide riparian zones, characterized by a cooler, wetter microclimate, can act as a buffer against wildfire and therefore as a refuge for fire- sensitive species (Pettit and Naiman 2007). Conversely, steep and highly-entrenched streams with narrow riparian zones are often characterized by more dense fuels than their adjacent upland forests (Van de Water and North 2011). In these cases, stream drainages can act as conduits for fire. For instance, there is evidence from montane ecosystems that riparian zones burn with equal or even greater frequency than upland forests (Van de

Water and North 2010, 2011), and that fire extent in riparian zones is highly correlated with fire extent in the upland (Arkle and Pilliod 2010).

Time since fire matters

In addition to varying with wildfire severity, responses of both organisms and ecosystem processes may differ in the short-term (days to months) versus longer-term

(years to decades) following fire. Immediate and short-term (days to one-year following wildfire) changes in riparian systems caused directly by wildfire may be short-lived, but their effects can persist over longer time periods. Direct effects of fire on soils and vegetation, for example, can influence the quantity and quality of water in these systems long after the fire (Shakesby and Doerr 2006). Interactions between wildfire and flooding generally result in patchy and temporally variable responses that can persist for months to decades (Pettit and Naiman 2007). For example, the first rain event following wildfire can be of particular importance in determining to what extent erosion and sedimentation occur. In the mid-term (often described as 2-10 years post-fire), there may be an increase in primary productivity both within the riparian zone (plants) and in the stream (benthic algae and macrophytes) due to increased light penetration. Conversely, stream reaches burned by low-severity fire may not differ from unburned streams in these respects within the short to mid-term (Jackson and Sullivan 2009, Malison and Baxter 2010a).

Over the long term (5-20 years), stream-riparian responses to high-severity wildfire are generally irregular and do not necessarily follow a direct succession. In addition, wildfire regimes are largely driven by climatic factors that vary greatly from year to year resulting in stochastic fire-return intervals (Agee 1993). Thus, the long-term consequences of

21 wildfire for stream-riparian organisms and ecosystem processes can be highly idiosyncratic and difficult to predict (see details below).

Spatial scale matters

Another important dimension determining stream-riparian ecosystem responses to wildfire severity is spatial scale. At the local scale (i.e., 101 - 102 m), fire can effect processes such as removal of canopy (Jackson and Sullivan 2009), mobilization of nutrients, erosion and sedimentation (Wondzell and King 2003), and channel stability

(Benda et al. 2003). At the catchment scale, fire can influence runoff timing and magnitude (Meyer and Pierce 2003), upland and riparian vegetation species composition and structure (Dwire and Kauffman 2003, Jackson and Sullivan 2009), local climate

(Rambo and North 2008), and habitat selection of organisms such as birds (Saab 1999), bats (Malison and Baxter 2010b, Buchalski et al. 2013), and fishes (Rieman and Clayton

1997, Dunham et al. 2003). Patches of open-canopy, large wood accumulation, sedimentation, and bank erosion that shift over time create habitat mosaics that can result in non-linear responses by aquatic and riparian organisms (Arkle et al. 2010). This highlights an important question of scale when it comes to assessing the impacts of wildfire of varying severity on stream-riparian habitats. Nearly all studies attempting to assess effects of wildfire on stream-riparian ecosystems have been focused at relatively small spatial scales and over relatively short time periods; understanding the cumulative effects of wildfire will require investigations of patterns that propagate through stream networks over longer periods of time (Benda et al. 2004, Burton 2005). The importance

22 of riparian areas as conduits for organisms and refugia for biodiversity (Sabo et al. 2005), combined with the upstream-to-downstream connectivity quintessential to stream ecosystems (Hynes 1975, Freeman et al. 2007), would suggest that riparian responses to wildfire have implications that extend from riverscapes to landscapes. Regardless, the lack of investigations across spatial scales points to an important uncertainty regarding our attempts at synthesis presented below. Studies are needed to address this gap in understanding the effects of wildfire severity.

Responses to a gradient of wildfire severity: evidence from the North American West

Responses to wildfire severity can be grouped into abiotic (physical and chemical) and biotic (individual organisms, populations, communities, and ecosystems).

Physical responses

In areas of low water volume, stream temperature can increase by several degrees during and immediately following (days to weeks) high-severity wildfire (Hitt 2003).

Over longer time periods (months to years), the loss of riparian vegetation and reorganization of the streambed due to post-fire shifts in channel geomorphology following severe wildfire can result in alterations to the heat budget of streams. Loss of shade and increased solar radiation results in higher stream temperatures (Dwire and

Kauffman 2003, Pettit and Naiman 2007). The magnitude of temperature change will be influenced by the severity of the fire, the total length of stream exposed, changes in riparian vegetation, and the degree of channel reorganization, with some streams showing

23 little response and others warming considerably (Royer and Minshall 1997, Dunham et al. 2007). Isaak et al. (2010) compiled a temperature database for a 2500-km river network in central Idaho to evaluate the effects of climate change and wildfire on stream temperatures and found that within wildfire perimeters, stream temperature increases were 2-3 times greater than basin averages, with radiation accounting for 50% of the warming.

Physical responses of streams and riparian zones such as alterations in hydrology and channel morphology tend to be persistent effects of wildfire, with immediate responses that can last decades past the fire event. For example, significant erosion and deposition of fine sediments in stream channels frequently follows high-severity fire

(Wondzell and King 2003). High-severity fire often consumes a significant portion of aboveground vegetation in the riparian zone and adjacent side slopes (Dwire and

Kauffman 2003). In addition, consumption of the litter layer and obstructions to overland water run-off like downed logs, conversion of organic material to small-particle ash, and development of hydrophobic soils (DeBano 2000, Doerr et al. 2003) can collectively contribute to reduced infiltration capacity of soils and the potential for increased overland flow, surface erosion, scouring of stream channels, and deposition of fine sediments

(Wondzell and King 2003, Shakesby and Doerr 2006, Vila-Escale et al. 2007). Under some circumstances, wildfire may be followed by debris flows; liquefied landslides that reorganize channels, export large wood, and can scour streambeds to bedrock (Miller et al. 2003, Wondzell and King 2003, May 2007). These and other physical disturbances that can accompany high-severity wildfire may extend and change the trajectory of post-

24 fire recovery of stream ecosystems. Whereas the local impacts of the wildfire-debris flow combination may lead to simplification of in-stream structure and morphology that may exert negative impacts on some stream organisms, this process also delivers sediment, wood, organic matter, and nutrients important to the complexity and character of downstream habitats (e.g. Benda et al. 2003, Harris et al. In revision-a).

Chemical responses

High-severity wildfires that consume the forest floor can considerably alter the magnitude and timing of overland flows, erosion, and solute delivery to streams

(Williams and Melack 1997, Seibert et al. 2010). Nutrients, contaminants, and organic compounds become concentrated after fire and can bind to fine sediments, thus increasing their transport into streams and elevating exposure to fish and aquatic invertebrates (Malmon et al. 2007). Partial combustion of riparian vegetation to ash that increases soil ammonium concentrations and results in increased stream nitrogen levels

(Wan et al. 2001) is also a common in-stream response to wildfire (Minshall et al. 1997,

Williams and Melack 1997, Bladon et al. 2008). Patterns of stream phosphorus concentrations following wildfire are less consistent, with evidence largely pointing to a brief (often returning to pre-fire conditions within a few weeks to a few months), but marked increase (e.g., Spencer and Hauer 1991, Hauer and Spencer 1998, Earl and Blinn

2003) or to no change (Minshall et al. 1997, Stephens et al. 2004). Overall, increases in nutrient delivery from the upland, combined with greater light penetration and higher

25 temperature, may prompt elevated in-stream primary productivity, with consequences for communities and food webs (Betts and Jones 2009: see "Food-web dynamics" below).

In contrast to physical responses, the chemical responses to wildfire generally have shorter-lived consequences (Minshall et al. 2003), largely because annual runoff often increases in the first couple years following fire (Moody and Martin 2001). For example, Hall and Lombardozzi (2008) found that the Hayman Fire, one of the largest wildfires in Colorado history (> 50-70% of the burn area classified as moderate- to high- severity fire) altered water temperature, dissolved oxygen concentrations, and concentrations of nitrate, phosphate, and mineral salts in stream water over the two-year post-burn period. Due to variability in climate, local topography, and burn characteristics, among other factors, chemical responses to moderate- and high-severity fires can be highly variable, with some streams returning to baseline conditions within weeks following fire (Earl and Blinn 2003) whereas other streams (or chemical constituents) show changes for multiple years (Hauer and Spencer 1998, Mast and Clow 2008). Effects on low severity fires on stream chemistry appear to be slight and typically do not persist beyond the first year (Stephens et al. 2004, Bêche et al. 2005). Though results of most studies suggest fire-driven shifts in chemistry are relatively ephemeral, such work has focused on the expected, pulsed delivery of materials from the land that follows fire. In contrast, and unlike research in the forested uplands (e.g., Smithwick et al. 2005, Koyama et al. 2010), there has been virtually no investigation of the mid- to long-term changes in biogeochemical processes that may accompany the more persistent changes in stream conditions or the biota that occupy riparian soils and streambed sediments.

Immediate effects on individuals

Responses to wildfire at the level of individual organisms are largely behavioral and physiological, and occur in immediate to short time periods following fire. Highly mobile animals, such as birds and mammals, can move to unimpacted areas away from high temperatures and smoke. Some terrestrial animals, including large ungulates like elk and bison, have been observed taking refuge in streams while wildfires are actively burning the upland (Allred et al. 2013). Though amphibians with in-stream life-cycle stages can lack the mobility to survive or move long distances in response to physical or chemical changes that might accompany wildfire (Gresswell 1999, Pilliod et al. 2003), some species can move to wet areas and/or burrow to avoid high-temperatures during wildfire. Although specific examples from the North American West are sparse, in

Australia, the anuran Hyperolius nitidulus can detect the sound of wildfire and seek refuge in wet areas (Grafe et al. 2002); American toads (Bufo americanus) were found partially buried in mud following a prescribed fire in Iowa, USA (Pilliod et al. 2003); and

Vogl (1973) discovered partially burned Leopard frogs (Rana sphenocephala) and Bull frogs (R. catesbiana) in a wetland following a fire in Florida. There is some evidence that lethal temperatures and/or changes in stream water chemistry during or shortly after wildfire can lead to mortality of fish and benthic invertebrates (Bozek and Young 1994,

Rinne 1996, Rieman and Clayton 1997, Howell 2006). However, even the direct, immediate effects of wildfire on benthic invertebrates are often negligible (Minshall

2003), and these impacts on fish can also be quite variable. In some cases, fish can be temporarily extirpated by the direct effects of high-severity fire (e.g., increased temperature, dissolved gases), especially in smaller streams (Dunham et al. 2003), but they often recover within weeks to months (Sestrich et al. 2011).

In-stream biotic response – populations and communities

The impact of wildfire on benthic invertebrate communities varies with fire severity and over time (Minshall 2003). Following the first large post-fire runoff, invertebrate richness may decline. Communities may recover to pre-fire conditions 1-2 years following wildfire, but a common pattern observed in many settings is that community composition shifts towards an increase in the relative abundance of disturbance-adapted taxa (Mihuc and Minshall 1995, Minshall 2003, Verkaik et al.

2013a). For example, in the short term following the Mortar Creek Fire in central Idaho, disturbance-adapted taxa were more dominant, but total taxa richness converged with reference streams towards the end of the 10-year study (Minshall et al. 2001a). In streams of the same region, Malison and Baxter (2010a) found that benthic insect assemblage composition continued to vary with fire severity five years following wildfire, with stream reaches that experienced high-severity fire having the greatest biomass of insects like midges (Chironomidae) and Baetid mayflies. Stream invertebrate communities may also shift following fire in terms of dominant feeding traits. Many disturbance-adapted taxa that flourish after fire are also feeding generalists (Mihuc and Minshall 1995) but, again, changes may be influenced by the state of riparian vegetation. For instance,

28 following the Jesisita Fire in southern California, shredders (which frequently rely on shredding leaves that fall into streams) were more abundant in streams draining unburned basins than those that burned but retained a riparian canopy, and were completely absent from basins where the riparian canopy was removed by fire (in these streams collector/filterer insects dominated) (Cooper et al. 2014). Therefore, although species composition and feeding-guild representation within benthic invertebrate communities varies along gradients of fire-severity and extent (Arkle et al. 2010), in the years to decades following fire the number of taxa may remain fairly consistent (Verkaik et al.

2013a).

Negative effects on fish may last for months to years following high-severity fire if elevated stream temperatures create stressful conditions, as can be the case for salmonids near the southern margin of their range (e.g., Beakes et al. 2014), or if stream reaches are also influenced by debris flows that may follow wildfire (Dunham et al.

2007). Fish adapted to cold-water habitats may be particularly sensitive to elevated temperatures that can occur during and in the few years following high-severity wildfires

(Sestrich et al. 2011) and can last ten years or more (Gresswell 1999, Isaak et al. 2010,

Sestrich et al. 2011). Post-fire high water temperatures have been linked to reduced density of salmonids in the American Southwest (Rinne 1996), Rocky Mountains (Isaak et al. 2010), and California (Beakes et al. 2014) because these fishes are especially dependent on cold water for spawning and juvenile rearing. For example, Sestrich et al.

(2011) found that pools in high-severity burned reaches of a western Montana stream were 2-6° C warmer in summer months in the year following fire and that native

29 westslope cutthroat trout (Oncorhynchus clarki lewisii) density was negatively correlated with percent of the catchment burned with moderate or high severity at one year post-fire.

This may be due to increased bioenergetic demand, as demonstrated in the study by

Beakes et al. (2014) who found that steelhead trout (O. mykiss) biomass was reduced in pools under canopy gaps one year after a fire in southern California. On the other hand, from one year to three years post-fire Sestrich et al. (2011) found higher proportions of moderate- and high-severity fire were associated with increases in populations of native fish.

Shifts in temperature associated with climate change may exacerbate short-term spikes in water temperature caused by fire, which may have added consequences for fish populations. Stream temperature increases from climate change were found to result in the reduction of spawning and rearing habitat for bull trout (Salvelinus confluentus) by 8 to 16% each decade in one study area in central Idaho; fire, at least temporarily, contributed to reductions in spawning/rearing habitat, though population levels were not recorded, so the extent to which, or whether, fire actually reduced bull trout in burned areas is unknown (Isaak et al. 2010). Because recovery of fish populations following fire may be influenced by recolonization of burned stream reaches from nearby unburned or low-severity burned reaches, drainage connectivity, ecosystem size, and timing of life- history events such as spawning, may interact with fire severity to influence population recovery (reviewed in Dunham et al. 2003). For example, endangered Gila trout (O. gilae) in New Mexico are especially vulnerable as they live in small isolated streams currently experiencing a frequent high-intensity fire regime (Propst et al. 1992).

Evidence regarding potential effects on the composition of fish assemblages is relatively sparse. Fishes like endangered salmonids in streams of the North American

West have relatively specific habitat needs, and if they are occurring in systems close to the edge of their range or those that are already degraded and fragmented, then they may be the most vulnerable to fire or associated disturbance and warming temperatures

(Dunham et al. 2003, Isaak et al. 2010, Beakes et al. 2014). On the other hand, direct investigations of fish composition responses have been few and results rather equivocal.

For instance, studies in Idaho and Montana have found little evidence of persistent, negative effects of even severe wildfire on salmonid fish assemblages (e.g., Neville et al.

2009, Sestrich et al. 2011). It has also been posited that effects of severe wildfire might facilitate invasions of nonnative fishes (Dunham et al. 2003). In streams of western

Montana, Sestrich et al. (2011) found no evidence of increases in abundance or invasion by eastern brook trout (Salvilinus fontinalis) after wildfire, but this hypothesis remains to be more widely tested.

Scouring flows that can result from high-severity wildfire can temporarily extirpate invertebrates, fish, and amphibians (Verkaik et al. 2013a). For example, Vieira et al. (2004) found that the first 100-year flood event following the 1996 Dome Fire in

New Mexico reduced benthic invertebrate density to near zero. However, within a year benthic invertebrate density recovered to pre-fire levels largely due to recolonization by those that disperse as larvae. Many fishes require relatively stable bed conditions with specific sediment class sizes for spawning; therefore, depending on the timing of fires, floods, and spawning, fish may be more or less affected by erosion and sedimentation.

Amphibian populations like the California newt (Taricha torosa) are similarly affected if preferred oviposition sites are filled with sediments (Gamradt and Kats 1997).

Over longer time scales (multiple years to decades) following wildfire, responses by stream organisms can be quite divergent, and this variation may be associated with the severity of wildfire and the trajectory of both stream and riparian habitat recovery. As described above, recovery within months to a few years is commonly observed among both invertebrate and fish populations, and debris flows in the early post-fire time period may later be associated with increased native fish populations (Sestrich et al. 2011).

However, over longer time-scales (multiple years) both groups may exhibit more interannual variability in stream reaches that have experienced high-severity wildfire than in those that burned with low-severity wildfire (Arkle and Pilliod 2010). Annual variability in populations is largely linked to floods and droughts (Robinson et al. 2000,

Arkle et al. 2010, Verkaik et al. 2013b). For example, in the Big Creek watershed of central Idaho, benthic invertebrate populations fluctuated annually with shifts correlated with a combination of sediments, large wood, riparian cover, and benthic organic matter along a gradient of fire-severity (Arkle et al. 2010), suggesting an interaction between fire severity and flooding in driving benthic invertebrate variability. On the other hand, in settings where drought is prevalent and accompanies wildfire, the impacts of stream drying on aquatic organisms may outweigh most variation associated with fire severity, as has been observed in Mediterranean and Australian stream ecosystems (Verkaik et al.

2013b, Verkaik et al. In revision). In any case, understanding the net consequences for stream organisms will likely require investigations that not only encompass different time

32 scales, but also responses at the scale of entire stream networks—studies that, thus far, are lacking.

Riparian community and ecosystem responses

Because of the pervasive influence of riparian plant composition and structure on a host of ecosystem responses, the influence of wildfire on riparian plant communities has received broad attention and highlights the importance of disturbance for driving composition and structure of stream-riparian communities. Riparian plants are highly adapted to disturbance (Naiman et al. 2005). In most cases this disturbance is flooding, and in certain biomes drought and fire (Pettit and Naiman 2007). Therefore, riparian plant species often possess distinct life-history traits such as stump sprouting, seed banks, and clonal regeneration that allow them to withstand fire or recover quickly following even severe wildfire. For example, plants in riparian forests often exhibit higher foliar moisture content that upland plants, even within the same species (Agee et al. 2002), which can result in patches of lower fire severity and lower plant mortality (Kauffman and Martin

1989, 1990). Tree species common to riparian areas in mountainous areas of the North

American West, such as ponderosa pine (Pinus ponderosa), western larch (Larix occidentalis), and coastal redwood (Sequoia sempervirens) have thick bark that protect them from mortality during low-intensity ground fires (Miller 2000). Low- and moderate- severity fire can stimulate clonal regeneration of quaking aspen (Populous tremuloides)

(Jones and DeByle 1985, Romme et al. 1997, Bartos and Campbell Jr. 1998), and aspen trees that are top-killed by high-severity fire are stimulated to produce numerous root

33 suckers (Schier 1973, Keyser et al. 2005). Many riparian shrubs, including alder (Alnus spp.), birch (Betula spp.), currant (Ribes spp.), rose (Rosa spp.), and snowberry

(Symphoricarpos spp.) sprout from stumps, root crowns, and belowground stems following fire (Adams et al. 1982, Stickney 1986, Miller 2000, Kobziar and McBride

2006) (Figure 3).

Species composition and taxonomic richness of riparian vegetation varies along a gradient of low- to high-severity wildfire and the trajectory of community response tends to not follow a predictable succession. One year following a prescribed fire in the Sierra

Nevada, Bêche et al. (2005) observed a reduction in species richness of riparian vegetation. However, five years following the Diamond Peak wildfire in central Idaho,

Jackson and Sullivan (2009) found species richness of riparian vegetation did not vary across stream reaches characterized as unburned, low-severity burned, and high-severity burned. Additionally, that study found riparian vegetation community composition did not differ between unburned and low-severity burned reaches, whereas high-severity burned reaches exhibited greater relative density of sun-loving species like blue elderberry (Sambucus cerulean) and red raspberry (Rubus ideaus). Moreover, herbaceous cover within high-severity burned reaches was dominated by invasive cheat grass

(Bromus tectorum), which has been associated with increased fire frequency and rate of spread (Mack and D'Antonio 1998).

The structure and composition of riparian vegetation is thought to be closely linked to recolonization dynamics of riparian invertebrates following high-severity wildfire, both in terms of species composition and when they recolonize. Bess et al.

(2002) found that the total number of riparian arthropod species was similar before (80) and nine months after (79) a high-severity fire in New Mexico. Out of the original 80, 30 species had not recovered, but 29 species that had not been recorded in the three years prior to the fire had appeared. Similarly, Jackson et al. (2012) found that five years after

35 wildfire the taxonomic composition of terrestrial invertebrates falling into streams differed between those that flowed through reaches burned by high-severity wildfire versus those that had experienced low-severity wildfire. However, these investigators observed the total number of taxa was consistent across burn types. Therefore, although species turnover appears to be common following high-severity wildfire, richness may remain similar.

Given that vegetation and habitat structure are critical factors that drive habitat selection in birds, wildfire in riparian zones can have substantial influence on bird communities as well. Some guilds, such as aerial insectivores, have been found to generally favor burned areas (Kotliar et al. 2002, Russell et al. 2009), potentially as a response to improved foraging conditions following reduced canopy cover, and increases in flying arthropods and emergent aquatic insects. Cavity-nesters (i.e., those that nest in sheltered chambers versus open-cup nests) are also thought to respond positively to wildfire, in part because the dense stands of snags (dead trees) created by wildfire provide important nesting sites (Saab and Powell 2005, Saab et al. 2009), although time since fire and fire severity influence these patterns. Some bird species specialize on habitats burned by high-severity fires (Hutto 2008) (See Text Box 1). In managed forests of the Sierra Nevada of California, riparian-associated birds increased in abundance < 6 years following low-severity prescribed fire (Bagne and Purcell 2011). On the other hand, ground-nesting red-faced warblers (Cardellina rubrifrons) and yellow-eyed juncos

(Junco phaeonotus) avoided nesting in riparian areas burned by low-severity fire one to two years following fire in southern Arizona (Kirkpatrick and Conway 2010). The timing

36 of fire may also be a particularly important factor for birds, as spring burns, for example, can interfere with breeding activities (Kruse and Piehl 1986). Thus, responses of bird communities to wildfire appear highly variable, benefitting some groups more than others.

Primary and secondary production

To understand the effects of wildfire on ecosystem processes, it is critical to consider its impacts on rates of both primary and secondary production. At present, inferences regarding these impacts must largely be drawn from studies that have measured indices of productivity, principally snap-shots of the biomass of producers like streambed algae or consumers like invertebrates. Increased algal biomass frequently occurs over the short- to mid-term after severe wildfire, likely due to increased light penetration to streams through canopy gaps combined with altered nutrient inputs

(Robinson et al. 1994, Minshall et al. 1997, Spencer et al. 2003). One year following the

Jesusita Fire in southern California, pools and riffles where the riparian canopy had been removed by fire exhibited 85% more cover by filamentous micro-algae than unburned or canopy intact reaches (Cooper et al. 2014). In the two years following the Diamond Peak

Fire in central Idaho, chlorophyll-a values and periphyton ash-free dry mass was significantly higher in burned streams (Rugenski and Minshall 2014). Similarly, in the only direct measure of primary productivity of which we are aware, during the year following a fire in Alaska, Betts and Jones (2009) observed rates of gross primary productivity of aquatic periphyton were double those in unburned sites. If elevated

37 temperatures or light inputs persist (as may occur if fires are severe and riparian canopies remain open), this pulse of aquatic productivity may endure as well, but this remains to be evaluated.

If wildfire leads to increased in-stream primary productivity, this may in turn contribute to higher rates of secondary production by benthic invertebrates, and such responses may be mediated by fire severity. For example, Malison and Baxter (2010b) found that benthic invertebrate biomass was fivefold greater in stream reaches that had been burned by high-severity wildfire five years prior compared with low-severity burned sites, and in the same study they reported that rates of emerging adult insects produced in reaches burned with high severity fire were three times higher than in those that were unburned or burned with low severity (Malison and Baxter 2010a). Although a similar pattern of elevated emergence was observed following fire in Washington (Mellon et al.

2008), the generality of these observations has not been evaluated and, remarkably, no study to date has measured annual rates of invertebrate production in response to wildfire.

Food-web dynamics

Food webs in streams and riparian zones are linked to one another via the bidirectional fluxes of materials and organisms. If increases in primary and secondary productivity do follow severe wildfire, this may have far-reaching consequences for organisms at higher trophic levels in stream-riparian food webs (what Malison and Baxter

2010b refer to as a "fire pulse"). For example, fish have been shown to selectively forage at the confluence of mainstem rivers and smaller tributaries that have been burned by

38 high-severity wildfire in the last five to ten years. Presumably this is due to greater export of benthic invertebrate prey originating from those tributaries (Koetsier et al.

2007) and a recent study shows that, indeed, tributaries disturbed by fire and associated debris flows export more invertebrate prey than those that were unburned, and fish exhibit preference for confluences within these disturbed streams (Harris et al. In revision-b). In addition, as emergent adults, stream insects are heavily relied upon as food resources for riparian consumers like birds, bats, and spiders (reviewed in Baxter et al.

2005). For example, in central Idaho, Malison and Baxter (2010a, 2010b) not only observed amplified emergence from sites that burned with high severity, they found that abundance of riparian web-building spiders from the family Tetragnathidae was two times higher in high-severity burned reaches. Conversely, Jackson and Sullivan (In press) found that Tetragnathidae density was not significantly different in stream reaches of

Yosemite National Park (Sierra Nevada Mountains, California) affected by low-severity fire compared to those affected by high-severity fire within the last 3-15 years, and this result was largely linked to differences in climate with sites that experienced more annual precipitation supporting greater density of riparian spiders.

From the riparian zone to the stream, high-severity wildfire can alter the magnitude, composition, and timing of inputs of leaf litter and terrestrial invertebrates.

Jackson et al. (2012) found that leaf litter inputs (dry weight) to streams five years after the Diamond Peak wildfire in central Idaho were 2-6 times greater in unburned reaches and 1.5-2 times greater in low-severity burned reaches compared to high-severity burned reaches where the riparian canopy was removed. In addition, inputs of terrestrial

39 invertebrates was as much as four-times greater to unburned reaches and two-times greater to low-severity burned reaches compared with high-severity burned reaches. The importance of terrestrial invertebrates as prey items for fish has been demonstrated in detail (Allan 1981, Wipfli 1997, Piccolo and Wipfli 2002, Allan et al. 2003, Carpenter et al. 2005), and high-severity wildfire may alter these subsidies. However, it appears that synchronized stimulation of in-stream primary and secondary productivity by high- severity wildfire combined with changes to habitat structure can result in a net neutral, or even beneficial, effect on in-stream and riparian consumers (See Text Box 2, Figure 4 on bats).

Text Box 2

The effects of wildfire on riparian habitat are expected to influence insectivorous bat distributions, foraging, and population dynamics. In forested systems, rivers create spatial gaps in dense forest vegetation, allowing echolocating bats to effectively forage directly over river channels, with comparatively low activity within or beneath the forest canopy

(Power et al. 2004, Ober and Hayes 2008). Riparian trees and snags also provide important roosting habitat for multiple riparian bat species (Brack 1983, Fleming et al.

2013). Fire-induced changes in riparian and bottomland vegetation structure therefore could have significant impacts on both bat habitat and energetics. For example, Buchalski et al. (2013) found that, in mixed-conifer forests of California, some bat species preferentially select moderate- and high-severity burned areas for foraging, likely

40 facilitated by reduced vegetation density and increased post-fire availability of prey and roosts.

Rivers and their adjacent riparian zones also provide important foraging habitat for insectivorous bats (Seidman and Zabel 2001, Russo and Jones 2003, Fukui et al.

2006), where aquatic insects that emerge from the stream as adults can comprise the majority of bat diets (Belwood and Fenton 1976, Swift et al. 1985). Higher aquatic insect availability is often implicated as the mechanism driving observations of higher rates of foraging activity within riverine landscapes as compared to upland habitats (Swift et al.

1985, Brigham et al. 1992). For example, Fukui et al. (2006) showed that bat activity along a Japanese stream significantly decreased after experimentally reducing aquatic insect emergence. In extremely arid climates, Hagen and Sabo (2012) found that seasonal river drying resulted in the disappearance of both aquatic insects and bats. Thus, because wildfire has the potential to profoundly alter aquatic insect emergence, terrestrial consumers such as riparian insectivorous bats (Sabo and Power 2002a, Paetzold et al.

2005) may also be affected. For instance, Malison and Baxter (2010b) observed the greatest number of bat echolocation calls at stream sites influenced 5 years prior by high- severity wildfire, suggesting that fires of different severity may have different effects on stream-riparian food webs via fire-induced changes in stream secondary productivity and subsequent aquatic insect emergence. Food availability has also been shown to be related to individual health, where it can mediate stress levels in bats with seasonally fluctuating resources (e.g., aerial insects) (Lewanzik et al. 2012). Thus, although the exact nature of

41 the responses may be species-specific, high-severity wildfire is expected to have strong impacts on riparian bats through both direct and indirect mechanisms.

The dynamics of linked stream-riparian food webs integrate wildfire impacts over both short and long time scales and across communities and ecosystems. The importance of riparian leaf litter, woody debris, and other plant material entering the stream from the riparian zone has a long history of study and appreciation (Vannote et al. 1980, Gregory

42 et al. 1991), and the importance of aquatic sources of energy moving into riparian zones

(e.g., via adult aquatic insect emergence, but also through flood pulses as and movements of other organisms) and terrestrial-aquatic feedback loops has received increasing attention (reviewed in Baxter et al. 2005). However, the role of wildfire with respect to these linkages is just starting to be described (Spencer et al. 2003, Malison and Baxter

2010b, Jackson et al. 2012).

Biodiversity, conservation, and management

Wildfire creates heterogeneous habitat conditions in space and time that may be important to the maintenance of native biodiversity and to the function of stream-riparian ecosystems. As we have summarized, whereas some organisms and processes may be negatively influenced by severe wildfire (at least on short time scales), many appear to be resilient over longer time periods. Indeed, stream-riparian ecosystems are often characterized as “dynamic mosaic[s] of spatial elements and ecological processes” (Ward et al. 2002). The creation and transformation of habitat patches and facilitation of ecological functions in streams and rivers is largely driven by disturbance: foremost flooding, but also drought, ecosystem engineers (e.g., beaver), and severe wildfire, as well as a host of human disturbances (e.g., dams, land-use, and climate change). The importance of flooding for the maintenance of ecological function and biodiversity of stream-riparian ecosystems has been demonstrated in detail and can result in high habitat and species turnover as compared to other ecosystems (Sullivan and Watzin 2009,

Tockner et al. 2010). For example, flooding creates and maintains a spatial mosaic of

43 habitats that in turn foster diverse webs of interacting species (Junk et al. 1989, Stanford et al. 2005, Bellmore et al. 2014), and homogenization of flow regimes by dams and other means has been shown to greatly reduce global biodiversity (Poff et al. 2007).

Similarly, because fire regimes are predictable over evolutionary timescales, it seems likely that alterations in the magnitude, frequency, timing, and extent of historic fire regimes will have consequences for stream-riparian biodiversity. Indeed, as has been the case for terrestrial ecosystems, a scientific consensus appears to be emerging regarding the importance and, in many instances, benefits, of wildfire to stream-riparian organisms and ecosystems, but at present there are at least two challenges to reaching such generalities that deserve consideration.

First, there remains a prevailing assumption that, whereas wildfire may have some ecological benefits, management for low-severity wildfire (or prescribed fire in its stead) should be preferred because it could represent a “goldilocks” condition, wherein such benefits might accrue while avoiding the perceived risks of high-severity fire. As we have summarized, the ecological effects (including various potential benefits) of high-severity wildfire in stream-riparian ecosystems do not appear to be mimicked by either low- severity wildfire or prescribed burning (e.g., Jackson and Sullivan 2009, Arkle and

Pilliod 2010, Malison and Baxter 2010b), which appears to call into question this assumption. On the other hand, interactions between severe wildfire and other sources of natural (flooding, drought, natural impoundments like debris jams) and anthropogenic disturbance (invasive species, post-fire logging, channel alteration and impoundments, introduction of nutrients and contaminants) might have cumulative or even exponential

44 effects on stream-riparian ecosystems, as have been described for other combinations of multiple stressors (Ormerod et al. 2010). For instance, connectivity in riverscapes

(Fausch et al. 2002) is likely important in mediating the local effects of severe wildfire on communities of native organisms; if waterways are disconnected by large patches of unsuitable habitat, organisms may not be able to redistribute following wildfire

(Gresswell 1999, Dunham et al. 2003). Studies are needed that explicitly evaluate how the sign and magnitude of responses to high-severity wildfire may differ with scale and in the context of other environmental stressors.

The second, and perhaps even more difficult, challenge is that, just as science is beginning to provide some understanding of the ways by which wildfire of varying severity may affect ecological function, the entire context for such relationships is being altered by a changing climate. In western North America, climate change has been linked to increases in wildfire frequency and extent (Whitlock et al. 2003, Whitlock 2004,

Westerling et al. 2006, Westerling et al. 2011a, Westerling et al. 2011b). This has been accompanied by changes in the trajectory of regional vegetation states (Allen and

Breshears 1998, van Mantgem et al. 2009), and some have hypothesized that many of the assumptions regarding resilience upon which existing fire ecology paradigms rest may now be poorly founded (see Davis et al. 2013 for review). This highlights the need to understand how wildfire characteristics and recovery patterns of terrestrial vegetation over time mediate responses of stream-riparian ecosystems. Yet, there have been proportionately few studies that explicitly evaluate how these might be influenced by its severity, and most investigations have focused on short-term responses, with far fewer

45 studies of mid- to longer term dynamics (i.e., > 2-3 yrs post-fire (Romme et al. 2011,

Rugenski and Minshall 2014). Such investigations will be needed to inform adaptive management of landscapes and riverscapes under changing climate.

A variety of management actions has been designed and used to mitigate the effects of severe wildfire on stream-riparian ecosystems, but the impact of these mitigation efforts is not always positive. For example, the use of prescribed fire as a tool to manage riparian ecosystem condition is increasing (Stone et al. 2010), but because prescribed fires typically differ from wildfires in severity, timing, frequency, and extent

(McIver et al. 2013), its influence on riparian and aquatic systems remains an open question (Boerner et al. 2008, Arkle and Pilliod 2010). In addition, methods utilized during fire suppression efforts can have negative effects on stream-riparian ecosystems.

For example, use of fire retardants around aquatic systems has led to mortality of aquatic organisms (Gaikowski et al. 1996, Buhl and Hamilton 2000, Gimenez et al. 2004) and is therefore banned by fire-fighting agencies, but construction of fire lines within drainages continues. In some cases, fire lines can facilitate the introduction of invasive species and be a significant source of chronic sediment delivery to streams following wildfires

(reviewed by Beschta et al. 2004, Karr et al. 2004).

Post-fire management has the potential to be more disruptive to stream-riparian ecosystems and have longer-lasting consequences than high-severity wildfire itself

(Beschta et al. 2004, Karr et al. 2004); therefore, any post-fire management that does not mitigate the effects of suppression activities should be avoided, including planting with non-native seeds, construction of debris dams, and post-fire logging. Debris dams are

46 often insufficient at ameliorating soil erosion and end up in stream channels following storms where they impede movement of organisms and disrupt flow. Mechanical disruption of soils, which often occurs as a result of post-fire logging, increases chronic erosion and deposition of fine sediments (McIver and Starr 2001, McIver and McNeil

2006), and soil compaction in forests can persist for 50-80 years (Quigley and Arbelbide

1997), which may exceed the duration of effects from high-severity wildfire. Even dead vegetation provides soil stability; snags are important habitat for riparian organisms, and large wood is a significant and ecologically important structural element of stream- riparian ecosystems (Gregory et al. 2003). Thus, post-fire logging may reduce the quality of stream-riparian habitat in multiple ways. Whereas post-fire management should be used with caution, pre-fire restoration of stream-riparian ecosystems might reduce potential negative effects of severe wildfire (Beschta et al. 2004); such efforts might include surfacing, stabilization, and removal of legacy roads, discouraging grazing in riparian zones, and restoration of fluvial connectivity.

Finally, as we have described in this chapter, the effects of wildfire on stream- riparian ecosystems operate over gradients of severity, space and time, and across levels of ecological organization. For example, although there are likely to be winners and losers at the individual and population level in the short term and over relatively small spatial scales, community- and ecosystem-level responses seem to be more neutral or positive, are longer lived, and tend to operate at relatively larger spatial scales. Therefore, management of stream-riparian ecosystems in landscapes that experience high-severity

47 fire will benefit from a holistic perspective that takes into account heterogeneous responses over space and time.

Conclusions

Wildfire plays an important role in shaping the structural and functional characteristics of stream- riparian ecosystems. Though these ecosystems represent a small portion of the total landscape, they are a disproportionately vital source of natural resources, and it is critical to understand how wildfire may influence them and the ecosystem services they provide. Riparian forests differ from upland forests in moisture regimes, microclimate, soils, topography, and vegetation, and can act as conduits of wildfire, burning with equal or even greater frequency than upland forests. Many of the diverse organisms associated with riparian corridors, ranging from fish and salamanders to birds and reptiles, are generally adapted to disturbance. Given that wildfire, in particular, has been a historic source of disturbance in many of these ecosystems, it is likely to be a key driver of biodiversity. Wildfire can impact streams and riparian zones at the individual, population, community, and ecosystem level. Its effects vary with fire severity, time since fire, and with the spatial scale of the fire. High-severity wildfire can have very different effects than low-severity wildfire, suggesting that a mosaic of fires of different severity may be necessary to maintain ecosystem function. Immediate and short- lived impacts such as increased stream temperatures can have negative impacts on individual organisms and populations, although evidence generally suggests that their

48 recovery is rapid, and there may be countervailing positive effects such as increased food availability for fish, bats, and birds in aquatic-riparian environments. Long-term impacts of wildfire are mediated by climate and can be irregular due to variation in site-specific physical characteristics. Additionally, the impact of wildfire may depend not only on fire severity, but also on the spatial scale (both total extent and patchiness) and timing of the fire, as well as on the degree of hydrologic connectivity. Streams and riparian zones are highly-connected pathways in landscapes; if this continuity is disrupted or other natural features (i.e., modified riparian zones, etc.) are impaired or lost due to human activities, this could lead to more detrimental effects of fire in these contexts. Due to the linked nature of stream-riparian ecosystems, and the disturbance-adapted organisms and food webs that characterize them, the role of wildfire in these ecosystems is likely essential to managing for biodiversity and conservation across the landscape. Further research is needed in the following areas to better understand and predict the effects of fire-severity in stream-riparian ecosystems and inform management: 1) Studies that investigate fire- severity effects over larger spatial scales and longer time periods that integrate fire extent, patchiness, and continuity; 2) Description of interaction effects between fire and other sources of both natural and anthropogenic disturbance; 3) Analysis of the ability of prescribed fires to emulate wildfires and provide ecosystem benefits; and 4) Longer term studies that integrate changes in fire regimes, vegetation, precipitation, and temperature due to climate change.

Literature Cited

Adams, D. E., R. C. Anderson, and S. L. Collins. 1982. Differential response of woody and herbaceous species to summer and winter burning in an Oklahoma grassland. The Southwest Naturalist 27:55-61. Agee, J. K. 1993. Fire ecology of the Pacific Northwest forests. Island Press, Washington, D.C. Agee, J. K., C. S. Wright, N. Williamson, and M. H. Huff. 2002. Foliar moisture content of Pacific Northwest vegetation and its relation to wildland fire behavior. Forest Ecology and Management 167:57-66. Allan, J. D. 1981. Determinants of diet of brook trout (Salvelinus fontinalis) in a mountain stream. Canadien Journal of Fishery and Aquatic Science 38:184-192. Allan, J. D., M. S. Wipfli, J. P. Caouette, A. Prussian, and J. Rogers. 2003. Influence of streamside vegetation on inputs of terrestrial invertebrates to salmonid food webs. Canadian Journal of Fisheries and Aquatic Sciences 60:309-320. Allen, C. D. and D. D. Breshears. 1998. Drought-induced shift of a forest-woodland ecotone: Rapid landscape response to climate variation. Proceedings of the National Academy of Sciences of the United States of America 95:14839-14842. Allred, B. W., S. D. Fuhlendorf, T. J. Hovick, R. D. Elmore, D. M. Engle, and A. Joern. 2013. Conservation implications of native and introduced ungulates in a changing climate. Global Change Biology 19:1875-1883. Arkle, R. S. and D. S. Pilliod. 2010. Prescribed fires as ecological surrogates for wildfires: A stream and riparian perspective. Forest Ecology and Management 259:893-903. Arkle, R. S., D. S. Pilliod, and K. Strickler. 2010. Fire, flow and dynamic equilibrium in stream macroinvertebrate communities. Freshwater Biology 55:299-314. Bagne, K. E. and K. L. Purcell. 2011. Short-Term Responses of birds to prescribed fire in fire-suppressed forests of California. Journal of Wildlife Management 75:1051- 1060. Bartos, D. L. and R. B. Campbell Jr. 1998. Decline of quaking aspen in the interior west, examples from Utah. Rangelands 20:17-24. Baxter, C. V., K. D. Fausch, and W. C. Saunders. 2005. Tangled webs: reciprocal flows of invertebrate prey link stream and riparian zones. Freshwater Biology 50:201- 220. Beakes, M. P., J. W. Moore, S. A. Hayes, and S. M. Sogard. 2014. Wildfire and the effects of shifting stream temperature on salmonids. Ecosphere 5. Bêche, L. A., S. L. Stephens, and V. H. Resh. 2005. Effects of prescribed fire on a Sierra Nevada (California, USA) stream and its riparian zone. Forest Ecology and Management 218:37-59. 50

Bellmore, J. R., C. V. Baxter, and P. J. Connolly. 2014. Spatial complexity reduces interaction strength in the meta-food web of a river floodplain mosaic. Ecology In press. Belwood, J. J. and M. B. Fenton. 1976. Variation in diet of Myotis lucifugus (Chiroptera- vespertilionidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie 54:1674-1678. Benda, L., D. Miller, P. Bigelow, and K. Andras. 2003. Effects of post-wildfire erosion on channel environments, Boise River, Idaho. Forest Ecology and Management 178:105-119. Benda, L., N. L. Poff, D. Miller, T. Dunne, G. Reeves, G. Pess, and M. Pollock. 2004. The network dynamic hypothesis: how channel networks structure riverine habitats. Bioscience 54:413-427. Beschta, R. L., J. J. Rhodes, J. B. Kauffman, R. E. Gresswell, G. W. Minshall, J. R. Karr, D. A. Perry, F. R. Hauer, and C. A. Frissell. 2004. Postfire management on forested public lands of the Western United States. Conservation Biology 18:957- 969. Bess, E. C., R. R. Parmenter, S. McCoy, and M. C. Molles. 2002. Responses of a riparian forest-floor arthropod community to wildfire in the middle Rio Grande Valley, New Mexico. Environmental Entomology 31:774-784. Betts, E. F. and J. B. Jones, Jr. 2009. Impact of wildfire on stream nutrient chemistry and ecosystem metabolism in boreal forest catchments of interior Alaska. Arctic Antarctic and Alpine Research 41:407-417. Bigelow, P. E., L. E. Benda, D. J. Miller, and K. M. Burnett. 2007. On debris flows, river networks, and the spatial structure of channel morphology. Forest Science 53:220- 238. Bisson, P. A., B. E. Rieman, C. Luce, P. F. Hessburg, D. C. Lee, J. L. Kershner, G. H. Reeves, and R. E. Gresswell. 2003. Fire and aquatic ecosystems of the western USA: current knowledge and key questions. Forest Ecology and Management 178:213-229. Bladon, K. D., U. Silins, M. J. Wagner, M. Stone, M. B. Emelko, C. A. Mendoza, K. J. Devito, and S. Boon. 2008. Wildfire impacts on nitrogen concentration and production from headwater streams in southern Alberta's Rocky Mountains. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 38:2359-2371. Blank, R. R., J. C. Chambers, and D. Zamudio. 2003. Restoring riparian corridors with fire: effects on soil and vegetation. Journal of Range Management 56:388-396. Boerner, R. E. J., J. Huang, and S. C. Hart. 2008. Impacts of fire and fire surrogate treatments on ecosystem nitrogen storage patterns: similarities and differences between forests of eastern and western North America. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 38:3056-3070. 51

Bozek, M. A. and M. K. Young. 1994. Fish mortality resulting from delayed-effects of fire in the greater Yellowstone ecosystem. Great Basin Naturalist 54:91-95. Brack, V. W. J. 1983. The nonhibernating ecology of bats in Indiana with emphasis on the endangered Indiana bat, Myotis sodalis. Purdue University, West Lafayette, Indiana, USA. Brigham, R. M., H. Aldridge, and R. L. Mackey. 1992. Variation in habitat use and prey selection by Yuma bats, Myotis yumanensis. Journal of Mammalogy 73:640-645. Buchalski, M. R., J. B. Fontaine, P. A. Heady, J. P. Hayes, and W. F. Frick. 2013. Bat response to differing fire severity in mixed-conifer forest California, USA. Plos One 8. Buhl, K. J. and S. J. Hamilton. 2000. Acute toxicity of fire-control chemicals, nitrogenous chemicals, and surfactants to rainbow trout. Transactions of the American Fisheries Society 129:408-418. Burton, T. A. 2005. Fish and stream habitat risks from uncharacteristic wildfire: observations from 17 years of fire-related disturbances on the Boise National Forest, Idaho. Forest Ecology and Management 211:140-149. Carpenter, S. R., J. J. Cole, M. L. Pace, M. Van de Bogert, D. L. Bade, D. Bastviken, C. M. Gille, J. R. Hodgson, J. F. Kitchell, and E. S. Kritzberg. 2005. Ecosystem subsidies: terrestrial support of aqautic food webs from 13C addition to contrasting lakes. 86 10:2737-2750 Carrier, W. D. and B. Czech. 1996. Threatened and endangered wildlife and livestock interactions. Pages 39-50 in P. R. Krausman, editor. Rangeland Wildlife. Society for Rangeland Management, Denver, CO. Cooper, S. D., H. M. Page, S. W. Wiseman, K. Klose, D. Bennet, T. Even, S. Sadro, C. E. Nelson, and T. L. Dudley. 2014. Physicochemical and biological responses of streams to wildfire severity in riparian zones. Freshwater Biology. e12523 Davis, J. M., C. V. Baxter, E. J. Rosi-Marshall, J. L. Pierce, and B. T. Crosby. 2013. Anticipating stream ecosystem responses to climate change: toward predictions that incorporate effects via land-water linkages. Ecosystems 16:909-922. DeBano, L. F. 2000. The role of fire and soil heating on water repellency in wildland environments: a review. Journal of Hydrology 231:195-206. Doerr, S. H., A. J. D. Ferreira, R. P. D. Walsh, R. A. Shakesby, G. Leighton-Boyce, and C. O. A. Coelho. 2003. Soil water repellency as a potential parameter in rainfall- runoff modelling: experimental evidence at point to catchment scales from Portugal. Hydrological Processes 17:363-377. Dunham, J. B., A. E. Rosenberger, C. H. Luce, and B. E. Rieman. 2007. Influences of wildfire and channel reorganization on spatial and temporal variation in stream temperature and distribution of fish and amphibians. Ecosystems 10:335-346.

Dunham, J. B., M. K. Young, R. E. Gresswell, and B. E. Rieman. 2003. Effects of fire on fish populations: landscape perspectives on peristance of native fishes and nonnative fish invasion. Forest Ecology and Management 178:183-196. Dwire, K. A. and J. B. Kauffman. 2003. Fire and riparian ecosystems in landscapes of the western USA. Forest Ecology and Management 178:61-73. Earl, S. R. and D. W. Blinn. 2003. Effects of wildfire ash on water chemistry and biota in South-Western USA streams. Freshwater Biology 48:1015-1030. Everett, R., R. Schellhaas, P. Ohlson, D. Spurbeck, and D. Keenum. 2003. Continuity in fire disturbance between riparian and adjacent sideslope Douglas-fir forest. Forest Ecology and Management 175:31-48. Fausch, K. D., C. E. Torgersen, C. V. Baxter, and H. W. Li. 2002. Landscapes to riverscapes: bridging the gap between research and conservation of stream fishes. Bioscience 52:483-498. Fleming, H. L., J. C. Jones, J. L. Belant, and D. M. Richardson. 2013. Multi-scale roost site selection by Rafinesque's Big-eared Bat (Corynorhinus rafinesquii) and Southeastern Myotis (Myotis austroriparius) in Mississippi. American Midland Naturalist 169:43-55. Freeman, M. C., C. M. Pringle, and C. R. Jackson. 2007. Hydrologic connectivity and the contribution of stream headwaters to ecological integrity at regional scales. Journal of the American Water Resources Association 43:5-14. Fukui, D., M. Murakami, S. Nakano, and T. Aoi. 2006. Effect of emergent aquatic insects on bat foraging in a riparian forest. Journal of Animal Ecology 75:1252-1258. Gaikowski, M. P., S. J. Hamilton, K. J. Buhl, S. F. McDonald, and C. H. Summers. 1996. Acute toxicity of three fire-retardant and two fire-suppressant foam formulations to the early life stages of rainbow trout (Oncorhynchus mykiss). Environmental Toxicology and Chemistry 15:1365-1374. Gamradt, S. C. and L. B. Kats. 1997. Impact of chaparral wildfire-induced sedimentation on oviposition of stream-breeding California newts (Taricha torosa). Oecologia 110:546-549. Gimenez, A., E. Pastor, L. Zarate, E. Planas, and J. Arnaldos. 2004. Long-term forest fire retardants: a review of quality, effectiveness, application and environmental considerations. International Journal of Wildland Fire 13:1-15. Grafe, T. U., S. Dobler, and K. E. Linsenmair. 2002. Frogs flee from the sound of fire. Proceedings of the Royal Society - Biological Sciences 269:999-1003. Gregory, S. V., A. M. Gurnell, and K. L. Boyer. 2003. The ecology and management of wood in world rivers. American Fisheries Society, Bethesda, Maryland. Gregory, S. V., F. J. Swanson, W. A. McKee, and K. W. Cummins. 1991. An ecosystem perspective of riparian zones; focus on links between land and water. Bioscience 41:540-552. 53

Gresswell, R. E. 1999. Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193-221. Hagen, E. M. and J. L. Sabo. 2012. Influence of river drying and insect availability on bat activity along the San Pedro River, Arizona (USA). Journal of Arid Environments 84:1-8. Hall, S. J. and D. Lombardozzi. 2008. Short-term effects of wildfire on montane stream ecosystems in the Southern Rocky Mountains: one and two years post-burn. Western North American Naturalist 68:453-462. Harris, H. E., C. V. Baxter, and J. M. Davis. In press. Debris flows mediate effects of wildfire on magnitude and composition of tributary subsidies to main-stem habitats. Freshwater Science. Harris, H. E., C. V. Baxter, and J. M. Davis. In revision. Wildfire and debris flows affect prey subsidies with implications for riparian and reverine predators. Freshwater Science. Hauer, F. R. and C. N. Spencer. 1998. Phosphorus and nitrogen dynamics in streams associated with wildfire: a study of immediate and longterm effects. International Journal of Wildland Fire 8:183-198. Hitt, N. P. 2003. Immediate effects of wildfire on stream temperature. Journal of Freshwater Ecology 18:171-173. Howell, P. J. 2006. Effects of wildfire and subsequent hydrologic events on fish distribution and abundance in tributaries of north fork John Day River. North American Journal of Fisheries Management 26:983-994. Hutto, R. L. 2008. The ecological importance of severe wildfire: some like it hot. Ecological Applications 18:1827-1834. Hynes, H. B. N. 1975. The stream and its valley. Verh. Internat. Verein. Limnol. 19:1-15. Isaak, D. J., C. H. Luce, B. E. Rieman, D. E. Nagel, E. E. Peterson, D. L. Horan, S. Parkes, and G. L. Chandler. 2010. Effects of climate change and wildfire on stream temperatures and salmonid thermal habitat in a mountain river network. Ecological Applications 20:1350-1371. Jackson, B. K. and S. M. P. Sullivan. 2009. Influence of fire severity on riparian vegetation heterogeneity in an Idaho, U.S.A. wilderness. Forest Ecology and Management 259:24-32. Jackson, B. K. and S. M. P. Sullivan. In press. Responses of riparian tetragnathid spiders to wildfire in forested ecosystems of the California Mediterranean climate region, USA. Freshwater Science. Jackson, B. K., S. M. P. Sullivan, and R. L. Malison. 2012. Wildfire severity mediates fluxes of plant material and terrestrial invertebrates to mountain streams. Forest Ecology and Management 278:27-34.

Jones, J. R. and N. V. DeByle. 1985. Fire. Pages 77-81 in N. V. DeByle and R. P. Winokur, editors. Aspen: ecology and management in the western United States. USDA Forest Service, Ogden, UT. Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood pulse concept in river- floodplain systems. Pages 110-127 in International Large River Symposium. Karr, J. R., J. J. Rhodes, G. W. Minshall, F. R. Hauer, R. L. Beschta, C. A. Frissell, and D. A. Perry. 2004. The effects of postfire salvage logging on aquatic ecosystems in the American West. Bioscience 54:1029-1033. Kauffman, J. B. and R. E. Martin. 1989. Fire behavior, fuel consumption, and forest-floor changes following prescribed understory fires in Sierra-Nevada mixed conifer forests. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 19:455-462. Kauffman, J. B. and R. E. Martin. 1990. Sprouting shrub response to different seasons and fuel consumtion levels of prescribed fire in Sierra-Nevada mixed conifer ecosystems. Forest Science 36:748-764. Keyser, T. L., F. W. Smith, and W. D. Shepperd. 2005. Trembling aspen response to a mixed-severity wildfire in the Black Hills, South Dakota, USA. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 35:2679-2684. Khorram, S. and E. F. Katibah. 1984. Vegetation and land cover mapping of the middle fork of the Feather River watershed using landsat digital data. Forest Science 30:248-258. Kirkpatrick, C. and C. J. Conway. 2010. Importance of montane riparian forest and influence of wildfire on nest-site selection of ground-nesting birds. Journal of Wildlife Management 74:729-738. Kobziar, L. N. and J. R. McBride. 2006. Wildfire burn patterns and riparian vegetation response along two Sierra Nevada streams. Forest Ecology and Management 222:254-265. Koetsier, P., Q. Tuckett, and J. White. 2007. Present effects of past wildfires on the diets of stream fish. Western North American Naturalist 67:429-438. Kotliar, N. B., S. J. Hejl, R. L. Hutto, V. A. Saab, C. P. Melcher, and M. E. McFadzen. 2002. Effects of fire and post-fire salvage logging on avian communities in conifer-dominated forests of the western United States. Studies in Avian Biology. 25, 49-64. Koyama, A., K. L. Kavanagh, and K. Stephan. 2010. Wildfire effects on soil gross nitrogen transformation rates in coniferous forests of central Idaho, USA. Ecosystems 13:1112-1126. Kruse, A. and J. Piehl. 1986. The impact of prescribed burning on groud-nesting birds. in G. Clambey and R. Pemble, editors. The prairie: past present and future. Proceedings of the 9th North American Prairie Conference, Moorhead, MN, 55

Center for Environmental Studies, Tri-College University, Fargo, North Dakota, USA. Lewanzik, D., D. H. Kelm, S. Greiner, M. Dehnhard, and C. C. Voigt. 2012. Ecological correlates of cortisol levels in two bat species with contrasting feeding habits. General and Comparative Endocrinology 177:104-112. Linder, K. A. and S. H. Anderson. 1998. Nesting habitat of Lewis' Woodpeckers in southeastern Wyoming. Journal of Field Ornithology 69:109-116. Mack, M. C. and C. M. D'Antonio. 1998. Impacts of biological invasions on disturbance regimes. Trends in Ecology and Evolution 13:195-198. Malison, R. L. and C. V. Baxter. 2010a. Effects of wildfire of varying severity on benthic stream insect assemblages and emergence. Journal of the North American Benthological Society 29:1324-1338. Malison, R. L. and C. V. Baxter. 2010b. The fire pulse: wildfire stimulates flux of aquatic prey to terrestrial habitats driving increases in riparian consumers. Canadian Journal of Fisheries and Aquatic Sciences 67:570-579. Malmon, D. V., S. L. Renean, D. Katzman, A. Lavine, and J. Lyman. 2007. Suspended sediment transport in an ephemeral stream following wildfire. Journal of Geophysics Research - Earth Surface 112:FO2006. Mast, M. A. and D. W. Clow. 2008. Effects of 2003 wildfires on stream chemistry in Glacier National Park, Montana. Hydrological Processes 22:5013-5023. May, C. 2007. Sediment and wood routing in steep headwater streams: an overview of geomorphic processes and their topographic signatures. Forest Science 53:119- 130. McIver, J. D. and R. McNeil. 2006. Soil disturbance and hill-slope sediment transport after logging of a severely burned site in northeastern Oregon. Western Journal of Applied Forestry 21:123-133. McIver, J. D. and L. Starr. 2001. A literature review on the environmental effects of postfire logging. Western Journal of Applied Forestry 16:159-168. McIver, J. D., S. L. Stephens, J. K. Agee, J. Barbour, R. E. J. Boerner, C. B. Edminster, K. L. Erickson, K. L. Farris, C. J. Fettig, C. E. Fiedler, S. Haase, S. C. Hart, J. E. Keeley, E. E. Knapp, J. F. Lehmkuhl, J. J. Moghaddas, W. Otrosina, K. W. Outcalt, D. W. Schwilk, C. N. Skinner, T. A. Waldrop, C. P. Weatherspoon, D. A. Yaussy, A. Youngblood, and S. Zack. 2013. Ecological effects of alternative fuel- reduction treatments: highlights of the National Fire and Fire Surrogate study (FFS). International Journal of Wildland Fire 22:63-82. Mellon, C. D., M. S. Wipfli, and J. L. Li. 2008. Effects of forest fire on headwater stream macroinvertebrate communities in eastern Washington, USA. Freshwater Biology 53:2331-2343.

Meyer, G. A. and J. L. Pierce. 2003. Climatic controls on fire-induced sediment pulses in Yellowstone National Park and central Idaho: a long-term perspective. Forest Ecology and Management 178:89-104. Mihuc, T. B. and G. W. Minshall. 1995. Trophic generalists vs trophic specialists - implications for food-web dynamics in post-fire streams. Ecology 76:2361-2372. Mihuc, T. B. and G. W. Minshall. 2005. The trophic basis of reference and post-fire stream food webs 10 years after wildfire in Yellowstone National Park. Aquatic Sciences 67:541-548. Miller, D., C. Luce, and L. Benda. 2003. Time, space, and episodicity of physical disturbance in streams. Forest Ecology and Management 178:121-140. Miller, M. 2000. Fire autecology. Pages 9-34 in J.K. Brown and J.K. Smith, editors. Wildland fire and ecosystems: effects of fire on flora. General Technical Report RMRS-GTR-412-vol2, USDA Forest Service, Ogden, UT. Minshall, G. F., J. T. Brock, D. A. Andrews, and C. T. Robinson. 2001a. Water quality, substratum and biotic responses of five central Idaho (USA) streams during the first year following the Mortar Creek fire. International Journal of Wildland Fire 10:185-199. Minshall, G. W. 2003. Responses of stream benthic macroinvertebrates to fire. Forest Ecology and Management 178:155-161. Minshall, G. W., K. E. Bowman, B. A. Rugenski, and C. Relyea. 2003. Monitoring of streams in the Payette National Forest 1988-2003: Big Creek and South Fork Salmon tributaries pre- and post-fire. USGS, Payette National Forest. Minshall, G. W., C. T. Robinson, and D. E. Lawrence. 1997. Postfire responses of lotic ecosystems in Yellowstone National Park, USA. Canadian Journal of Fisheries and Aquatic Sciences 54:2509-2525. Minshall, G. W., C. T. Robinson, D. E. Lawrence, D. A. Andrews, and J. T. Brock. 2001b. Benthic macroinvertebrate assemblages in five central Idaho (USA) streams over a 10-year period following disturbance by wildfire. International Journal of Wildland Fire 10:201-213. Moody, J. A. and D. A. Martin. 2001. Post-fire, rainfall intensity-peak discharge relations for three mountainous watersheds in the western USA. Hydrological Processes 15:2981-2993. Naiman, R. J., H. Decamps, and M. E. McClain. 2005. Riparia: ecology, conservation, and management of streamside communities. Elsevier Academic Press, Burlington, MA. Neville, H., J. Dunham, A. Rosenberger, J. Umek, and B. Nelson. 2009. Influences of wildfire, habitat size, and connectivity on trout in headwater streams revealed by patterns of genetic diversity. Transactions of the American Fisheries Society 138:1314-1327. 57

Ober, H. K. and J. P. Hayes. 2008. Influence of vegetation on bat use of riparian areas at multiple spatial scales. Journal of Wildlife Management 72:396-404. Ormerod, S. J., M. Dobson, A. G. Hildrew, and C. R. Townsend. 2010. Multiple stressors in freshwater ecosystems. Freshwater Biology 55:1-4. Paetzold, A. C., C. J. Schubert, and K. Tockner. 2005. Aquatic-terrestrial linkages along a braided river: riparian arthropods feeding on aquatic insects. Ecosystems 8:748- 758. Pettit, N. E. and R. J. Naiman. 2007. Fire in the riparian zone: characteristics and ecological consequences. Ecosystems 10:673-688. Piccolo, J. J. and M. S. Wipfli. 2002. Does red alder (Alnus rubra) in upland riparian forests elevate macroinvertebrate and detritus export from headwater streams to downstream habitats in southeastern Alaska? Canadian Journal of Fisheries and Aquatic Sciences 59:503-513. Pilliod, D. S., R. B. Bury, E. J. Hyde, C. A. Pearl, and P. S. Corn. 2003. Fire and amphibians in North America. Forest Ecology and Management 178:163-181. Poff, N. L., J. D. Olden, D. M. Merritt, and D. M. Pepin. 2007. Homogenization of regional river dynamics by dams and global biodiversity implications. Proceedings of the National Academy of Sciences of the United States of America 104:5732-5737. Postel, S. and S. Carpenter. 1997. Freshwater ecosystem services. Pages 195-214 in G. Daily, editor. Nature's services: societal dependence on natural ecosystems. Island Press. Power, M. E. and W. E. Dietrich. 2002. Food webs in river networks. Ecological Research 17:451-471. Power, M. E., W. E. Rainey, M. S. Parker, J. L. Sabo, A. Smyth, S. Khandwala, J. C. Finlay, F. C. McNeely, K. Marsee, and C. Anderson. 2004. River to watershed subsidies in an old-growth conifer forest. Pages 217-240 in G. A. Polis, M. E. Power, and G. R. Huxel, editors. Food webs at the landscape level. The University of Chicago Press, Chicago, IL, USA. Propst, D. L., J. A. Stefferud, and P. R. Turner. 1992. Conservation and status of Gila trout, Oncorhynchus gilae. Southwestern Naturalist 37:117-125. Pyne, S. J. 1997. America's fires: management on wildlands and forests. Forest History Society, Durham, North Carolina. Pyne, S. J. 2004. Tending Fire: coping with America's wildland fires. Washington Island Press. Quigley, T. M. and S. J. Arbelbide. 1997. An assessment of ecosystem components in the Interior Columbia Basin and portions of the Klamath and Great Basins. General Technical Report PNW-GTR-405. USDA Forest Service, Portland, Oregon, USA.

Rambo, T. R. and M. P. North. 2008. Spatial and temporal variability of canopy microclimate in a Sierra Nevada riparian forest. Northwest Science 82:259-268. Resh, V. H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W. Li, G. W. Minshall, S. R. Reice, A. L. Sheldon, J. B. Wallace, and R. C. Wissmar. 1988. The role of disturbance in stream ecology. Journal of the North American Benthological Society 7:433-455. Rieman, B. E. and J. Clayton. 1997. Wildfire and native fish: issues of forest health and conservation of sensitive species. Fisheries 22:6-15. Rieman, B. E., D. Lee, G. Chandler, and D. Meyer. 1997. Does wildfire threaten extinctions for salmonids? Responses or redband trout and bull trout following recent large fires on the Boise National Forest. Pages 47-57 in Symposium on fire effects on threatened and endangered species and habitats. International Association of Wildland Fire, Fairfield. Rinne, J. N. 1996. Short-term effects of wildfire on fishes and aquatic macroinvertebrates in the southwestern United States. North American Journal of Fisheries Management 16:653-658. Robinson, C. T., G. W. Minshall, and T. V. Royer. 2000. Inter-annual patterns in macroinvertebrate communities of wilderness streams in Idaho, USA. Hydrobiologia 421:187-198. Robinson, C. T., S. R. Rushforth, and G. W. Minshall. 1994. Diatom asssemblages of streams influenced by wildfire. Journal of Phycology 30:209-216. Romme, W. H., M. S. Boyce, R. Gresswell, E. H. Merrill, G. W. Minshall, C. Whitlock, and M. G. Turner. 2011. Twenty years after the 1988 Yellowstone fires: lessons about disturbance and ecosystems. Ecosystems 14:1196-1215. Romme, W. H., M. G. Turner, R. H. Gardner, W. W. Hargrove, G. A. Tuskan, D. G. Despain, and R. A. Renkin. 1997. A rare episode of sexual reproduction in Aspen (Populus tremuloides Michx) following the 1988 Yellowstone fires. Natural Areas Journal 17:17-25. Royer, T. V. and G. W. Minshall. 1997. Temperature patterns in small streams following wildfire. Archiv Fur Hydrobiologie 140:237-242. Rugenski, A. T. and G. W. Minshall. 2014. Climate-moderated responses to wildfire by macroinvertebrates and basal food resources in montane wilderness streams. Ecosphere 5. Russell, R. E., J. A. Royle, V. A. Saab, J. F. Lehmkuhl, W. M. Block, and J. R. Sauer. 2009. Modeling the effects of environmental disturbance on wildlife communities: avian responses to prescribed fire. Ecological Applications 19:1253-1263.

Russo, D. and G. Jones. 2003. Use of foraging habitats by bats in a Mediterranean area determined by acoustic surveys: conservation implications. Ecography 26:197- 209. Saab, V. 1999. Importance of spatial scale to habitat use by breeding birds in riparian forests: a hierarchical analysis. Ecological Applications 9:135-151. Saab, V. A. and H. D. W. Powell. 2005. Fire and avian ecology in North America: process influencing pattern. Studies in Avian Biology:1-13. Saab, V. A., R. E. Russell, and J. G. Dudley. 2009. Nest-site selection by cavity-nesting birds in relation to postfire salvage logging. Forest Ecology and Management 257:151-159. Sabo, J. L. and M. E. Power. 2002. Numerical response of lizards to aquatic insects and short-term consequences for terrestrial prey. Ecology 83:3023-3036. Sabo, J. L., R. A. Sponseller, M. Dixon, K. Gade, T. Harms, J. Heffernan, A. Jani, G. Katz, C. U. Soykan, J. Watts, and J. Welter. 2005. Riparian zones increase regional species richness by harboring different, not more, species. Ecology 86:56-62. Schier, G. A. 1973. Origin and development of aspen root suckers. Canadian Journal of Forest Research 3:45-53. Seibert, J., J. J. McDonnell, and R. D. Woodsmith. 2010. Effects of wildfire on catchment runoff response: a modelling approach to detect changes in snow-dominated forested catchments. Hydrology Research 41:378-390. Seidman, V. M. and C. J. Zabel. 2001. Bat activity along intermittent streams in Northwestern California. Journal of Mammalogy 82:738-747. Sestrich, C. M., T. E. McMahon, and M. K. Young. 2011. Influence of fire on native and nonnative salmonid populations and habitat in a western Montana basin. Transactions of the American Fisheries Society 140:136-146. Shakesby, R. A. and S. H. Doerr. 2006. Wildfire as a hydrological and geomorphological agent. Earth-Science Reviews 74:269-307. Smithwick, E. A. H., M. G. Turner, M. C. Mack, and F. S. Chapin. 2005. Postfire soil N cycling in northern conifer forests affected by severe, stand-replacing wildfires. Ecosystems 8:163-181. Spencer, C. N., K. O. Gabel, and F. R. Hauer. 2003. Wildfire effects on stream food webs and nutrient dynamics in Glacier National Park, USA. Forest Ecology and Management 178:141-153. Spencer, C. N. and F. R. Hauer. 1991. Phosphorus and nitrogen dynamics in streams during a wildfire. Journal of the North American Benthological Society 10:24-30. Stanford, J. A., M. S. Lorang, and F. R. Hauer. 2005. The shifting habitat mosaic of river ecosystems. Verh. Internat. Verein. Limnol. 29:123-136.

Stephens, S. L., t. Meixner, M. Poth, B. McGurk, and D. Payne. 2004. Prescribed fire, soils, and stream water chemistry in a watershed in the Lake Tahoe Basin, California. International Journal of Wildland Fire 13:27-35. Stickney, P. F. 1986. First decade plant succession following the sundance forest fire, Northern Idaho. USDA Forest Service. General Technical Report GTR-INT- 197:26 Stone, K. R., D. S. Pilliod, K. A. Dwire, C. C. Rhoades, S. P. Wollrab, and M. K. Young. 2010. Fuel reduction management practices in riparian areas of the western USA. Environmental Management 46:91-100. Sullivan, S. M. P. and M. C. Watzin. 2009. Stream-floodplain connectivity and fish assemblage diversity in the Champlain Valley, Vermont, U.S.A. Journal of Fish Biology 74:1394-1418. Sweeney, B. W., T. L. Bott, J. K. Jackson, L. A. Kaplan, J. D. Newbold, L. J. Standley, W. C. Hession, and R. J. Horwitz. 2004. Riparian deforestation, stream narrowing, and loss of stream ecosystem services. Proceedings of the National Academy of Sciences of the United States of America 101:14132-14137. Swift, S. M., P. A. Racey, and M. I. Avery. 1985. Feeding ecology of Pipestrellus pipistrellus (Chiroptera: Vespertilionidae) during pregnancy and lactation. II. Diet. Journal of Animal Ecology 54:217-225. Tockner, K., M. S. Lorang, and J. A. Stanford. 2010. River floodplains are model ecosystems to test general hydrogeomorphic and ecological concepts. River Research and Applications 26:76-86. Van de Water, K. and M. North. 2010. Fire history of coniferous riparian forests in the Sierra Nevada. Forest Ecology and Management 260:384-395. Van de Water, K. and M. North. 2011. Stand structure, fuel loads, and fire behavior in riparian and upland forests, Sierra Nevada Mountains, USA; a comparison of current and reconstructed conditions. Forest Ecology and Management 262:215- 228. Van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z. Fule, M. E. Harmon, A. J. Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen. 2009. Widespread increase of tree mortality rates in the western United States. Science 323:521-524. Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Canadian Journal of Fishery and Aquatic Science 37:130-137. Verkaik, I., M. Rieradevall, S. D. Cooper, J. M. Melack, T. L. Dudley, and N. Prat. 2013a. Fire as a disturbance in Mediterranean climate streams. Hydrobiologia 719:353-382.

Verkaik, I., M. Vila-Escale, M. Rieradevall, C. V. Baxter, P. S. Lake, G. F. Minshall, P. Reich, and N. Prat. In press. Stream macroinvertebrate community responses to fire: are they the same in different fire-prone biogeographic regions? Freshwater Science. Verkaik, I., M. Vila-Escale, M. Rieradevall, and N. Prat. 2013b. Seasonal drought plays a stronger role than wildfire in shaping macroinvertebrate communities of Mediterranean streams. International Review of Hydrobiology 98:271-283. Vieira, N. K., W. H. Clements, L. S. Guevara, and B. F. Jacobs. 2004. Resistance and resilience of stream insect communities to repeated hydrologic disturbances after a wildfire. Freshwater Biology 49:1243-1259. Vierling, K. T. and V. Saab. 2004. Spatial scaling of woodpecker habitat selection in post-fire environments. Ecological Society of America Annual Meeting Abstracts 89:523-523. Vila-Escale, M., T. Vegas-Vilarrubia, and N. Prat. 2007. Release of polycyclic aromatic compounds into a Mediterranean creek (Catalonia, NE Spain) after a forest fire. Water Research 41:2171-2179. Vogl, R. J. 1973. Effects of fire on plants and animals of a Florida wetland. American Midland Naturalist 89:334-347. Wallace, J. B., S. L. Eggert, J. L. Meyer, and J. R. Webster. 1997. Multiple trophic levels of a forested stream linked to terrestrial litter inputs. Science 277:102-104. Wan, S., D. Hui, and Y. Luo. 2001. Fire effects on nitrogen pools and dynamics in terrestrial ecosystems: a meta-analysis. Ecological Applications 11:1349-1365. Ward, J. V., K. Tockner, D. B. Arscott, and C. Claret. 2002. Riverine landscape diversity. Freshwater Biology 47:517-539. Westerling, A. L., B. P. Bryant, H. K. Preisler, T. P. Holmes, H. G. Hidalgo, T. Das, and S. R. Shrestha. 2011a. Climate change and growth scenarios for California wildfire. Climatic Change 109:445-463. Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and earlier spring increase western US forest wildfire activity. Science 313:940-943. Westerling, A. L., M. G. Turner, E. A. H. Smithwick, W. H. Romme, and M. G. Ryan. 2011b. Continued warming could transform Greater Yellowstone fire regimes by mid-21st century. Proceedings of the National Academy of Sciences of the United States of America 108:13165-13170. Whitlock, C. 2004. Forests, fires and climate. Nature 432:28-29. Whitlock, C., S. L. Shafer, and J. Marlon. 2003. The role of climate and vegetation change in shaping past and future fire regimes in the northwestern US and the implications for ecosystem management. Forest Ecology and Management 178:5- 21.

Williams, M. R. and J. M. Melack. 1997. Effects of prescribed burning and drought on the solute chemistry of mixed-conifer forest streams of the Sierra Nevada, California. Biogeochemistry 39:225-253. Wipfli, M. S. 1997. Terrestrial invertebrates as salmonid prey and nitrogen sources in streams: contrasting old-growth and young-growth riparian forests in southeastern Alaska, USA. Canadian Journal of Fisheries and Aquatic Sciences 54:1259-1269. Wondzell, S. M. and J. G. King. 2003. Postfire erosional processes in the Pacific Northwest and Rocky Mountain regions. Forest Ecology and Management 178:75-87.

Chapter 2: Antecedent and recent wildfire severity in forested ecosystems of the Sierra Nevada, California, USA do not result in heterogeneous patterns in riparian vegetation and stream geomorphology

Breeanne K. Jackson; S. Mažeika P. Sullivan

Abstract: Stream geomorphology and riparian vegetation are structural components of stream-riparian ecosystems that contribute to biodiversity and ecosystem function. The strong effects of wildfire severity on structural components of stream systems in temperate zones might suggest equally strong influences on stream-riparian structure in

Mediterranean-type systems. However, at 12 stream reaches paired by fire-severity (one high-severity burned, one low-severity burned) in the central Sierra Nevada range,

California, USA, we found no significant differences in riparian plant community structure and composition or stream geomorphology. In addition, we observed minimal changes in riparian vegetation and stream geomorphology in the first summer following the extensive and severe Rim Fire. In contrast, fire-frequency at the catchment scale was significantly correlated with fluvial geomorphic metrics. This result, combined with the long period of time since fire at some study reaches (3-15 years), indicates that effects of fire severity on structural responses of stream-riparian ecosystems are not consistent over spatial extent and time. Further, inconsistent effects of wildfire over space and time may be related to high interannual variability in precipitation and seasonal dry-wet cycles characteristic of Mediterranean-type ecosystems.

Introduction

The influence of wildfire as a disturbance agent in stream-riparian ecosystems has been the focus of increasing attention (Gresswell 1999, Bisson et al. 2003, Verkaik et al.

2013a). In particular, whereas hydrologic disturbance plays an important role in shaping in-stream habitat structure (Gregory et al. 1991, Poff 1992, Stanley et al. 2010) as well as riparian and floodplain habitat patches (Johansson and Nilsson 2002, Stanford et al.

2005), fire may act as a terrestrial disturbance regime affecting both in-stream and riparian habitat structure (Benda et al. 2003, Rice et al. 2012). The effects of wildfire disturbance on stream-riparian systems might be expected to be particularly evident in

Mediterranean-type systems, where relatively mild and wet winters that promote the growth of plant biomass are followed by long, hot, and dry summers that support frequent wildfire (Halsey 2008).

Physical responses of stream-riparian ecosystems to wildfire include alterations in hydrology, channel morphology, and sediment delivery to streams (Benda et al. 2003,

Wondzell and King 2003); these responses typically occur immediately following a fire event but can last decades after wildfires occur (Sala et al. 1994, Shakesby and Doerr

2006, Mayor et al. 2007). Consumption of riparian and sideslope vegetation, leaf litter, and obstruction of overland water run-off (e.g., downed logs) combined with conversion of soil organic matter to small-particle ash, and, in some cases, development of hydrophic soils (DeBano 2000, Doerr et al. 2003) can collectively contribute to reduced infiltration capacity of soils and increase overland flow, surface erosion, scouring of stream channels, and deposition of fine sediments (Wondzell and King 2003, Shakesby and

Doerr 2006, Vila-Escale et al. 2007). The extent to which wildfire-induced changes in stream-riparian physical condition occurs is highly dependent on fire severity (i.e., low- severity fires do not often lead to changes in erosion and sedimentation; Bêche et al.

2005), and the timing and magnitude of post-fire storms (Swanson 1981, Shakesby

2011). In temperate systems, stream flooding typically follows predictable snow-melt and spring runoff, but in Mediterranean regions wet storms can immediately follow autumn wildfires resulting in less predictable interactions between fires and floods (Scott 1993,

DeBano 2000, Verkaik et al. 2013a).

Vegetation composition often reflects hydrologic and stream geomorphic processes (Polvi 2011), suggesting an important interaction between the two (Shaw 2008,

Bertoldi et al. 2011). The interaction between stream hydrogeomorphology and riparian vegetation may be more variable in Mediterranean systems compared to temperate systems. Although resistance to wildfire and rapid recovery of vegetation might suggest that erosion and sedimentation would be reduced in Mediterranean forests, wildfire in these regions frequently occurs in steep valleys with highly erodible soils (Shakesby

2011). In addition, high-intensity, post-fire, fall flooding typical of Mediterranean systems can lead to elevated sediment delivery to streams (Swanson 1981, Sala et al.

1994, Andreu et al. 2001) and increase the likelihood of landslides, debris flows, and rill erosion (Barro and Conard 1991). Because riparian vegetation plays an important role in attenuating bank erosion and intercepting sediments (Kobziar and McBride 2006, Pettit and Naiman 2007a), the structure and composition of riparian vegetation may be important for mediating effects on stream hydrogeomophology in the years to decades

66 following wildfire of varying severity. In contrast, persistent scouring flows and bank erosion may forestall recovery of riparian vegetation (Davis et al. 1989, Bendix and

Cowell 2010a).

Riparian forests in Mediterranean-type climates have been less studied than in temperate biomes and likely respond differently to wildfire severity than in temperate climates (Verkaik et al. 2013a). Unlike temperate forests where wildfire extent and severity can be largely driven by fuel continuity, structure, and loading, wildfires in

Mediterranean-type regions are more often governed by weather (i.e., drought and high winds; Pausas and Paula 2012), which may result in more heterogeneous post-fire conditions (i.e., patches of low-severity burned and unburned areas with reduced plant mortality) in riparian zones due to lower temperatures, higher humidity, and higher foliar moisture content than in the adjacent uplands. Mediterranean-type forests also differ from temperate forests in recovery times following wildfire (Dwire and Kauffman 2003).

Whereas the conifer canopy in temperate riparian forests may take centuries following severe wildfire to be replaced (Minshall et al. 1989), due to year-round growing conditions and a proliferation of fire-adapted species, pre-fire canopy levels may return in as little as 10-20 years in Mediterranean-type riparian forests (McMichael et al. 2004).

We examined the influence of low- and high-severity wildfire (i.e., low-severity wildfire burned only in the understory and the canopy remained intact; high-severity wildfire removed the riparian canopy) occurring over the last two decades on two structural elements – fluvial geomorphology and riparian vegetation – of linked stream- riparian ecosystems in Yosemite National Park (YNP) of the central Sierra Nevada,

California, USA. In regard to riparian vegetation, we hypothesized that riparian vegetation community composition would be significantly different between high- and low-severity burned stream reaches. We predicted that (1) Removal of the conifer canopy by high-severity wildfire and concurrent increases in light penetration would result in rapid recovery of riparian shrub and herb species resilient to wildfire and greater cover in these layers compared to shrub and herb layers in low-severity burned reaches; (2)

Community composition of riparian plants in low-severity burned reaches (where the canopy was retained) would be significantly different than in high-severity burned reaches such that shade-tolerant species would dominate and species richness would be higher due to greater habitat heterogeneity resulting from small canopy gaps and incomplete burning of the understory; and (3) High-severity wildfire, unstable banks, and elevated sedimentation would provide an early-succession environment suitable for colonization by invasive plant species. Concerning fluvial geomorphology, we hypothesized that stream channel geometry and the size and distribution of sediments would be significantly different between high-severity and low-severity burned stream reaches and that these differences would be correlated with the composition of riparian vegetation. We predicted that (1) Due to a loss of vegetation and increased soil erosion following high-severity wildfire, streams would exhibit a decrease in D50 (mean sediment size) and an increase in embeddedness (degree to which cobbles are surrounded by fine particles); (2) High-severity burned reaches would be characterized by increased entrenchment, incision, and channel-widening (as measured by width-to-depth ratio) compared to low-severity wildfire streams primarily as a result of changes in the balance

68 between sediment (supply, size) and flow (discharge, slope) dynamics (Lane 1955); and

(3) Characteristics of channel geometry (i.e., degree of channel widening, channel gradient (i.e., slope, entrenchment, incision), sediment size distribution, and embeddedness would be correlated with community composition of riparian vegetation.

Although our study focused on effects of fire severity at the local-scale (i.e., reach), we also explored potential influences of catchment-scale fire variability on stream geomorphic characteristics given its demonstrated importance to fluvial geomorphic processes (Arkle et al. 2010, Shakesby 2011).

Methods

Phase 1: Paired design (2011 and 2012)

Study area

Yosemite National Park of California’s central Sierra Nevada range covers 3,027 km2 with 95% designated as wilderness. The regional climate is Mediterranean and the park typically receives 94.5 cm of precipitation annually of which 73.7 cm falls as snow

(Western Regional Climate Center 2012). Although spring runoff is common, this region is heavily influenced by the El Niño Southern Oscillation (ENSO) cycle (DeFlorio et al.

2013). Therefore, precipitation patterns can be highly variable, characterized by fall and winter rain-on-snow events and floods as well as periods of sustained drought. Two

National Wild and Scenic Rivers (the Tuolumne and Merced Rivers) begin in the Park and flow westward through glacial valleys and join the San Joaquin River in the Central

Valley. The Merced River drains 4,470 km2 and has an average discharge of 34 m3 s-1 at

69 its mouth, while the Tuolumne basin encompasses 5,076 km2 and discharges 70 m3 s-1 on average (USGS 2012). The headwaters of the Merced River begin in the Clark Range at about 2,413 m above sea level and the Tuolumne begins at the confluence of the Dana and Lyell Forks in Tuolumne Meadows at 2,616 m above sea level.

The average fire-return interval for riparian forests of the Sierra Nevada is 16.6 to

30.0 years, which is comparable to the average for upland forests in the same region (16.9 to 27.8 years) (Van de Water and North 2010). In Yosemite National Park, fire frequency and annual extent increased markedly in the 1980s after wildfires were allowed to proliferate in the Park following a long history of fire suppression (Figure 5).

Common dominant vegetation zones include Yellow Pine belt at 600-2,000 m,

Lodgepole Pine and Red Fir forest at 2,000-2,500 m, and Subalpine belt at 2,250-3,000 m. Where fires have occurred, Mountain Chaparral can dominate these vegetation zones.

Dominant upland species include Ponderosa pine (Pinus ponderosa), sugar pine (Pinus lambertiana), red fir (Abies concolor), lodgepole pine (Pinus contorta murrayana), incense cedar (Libocedrus decurrens), chinquapin (Castanopsis sempervirens), and buck brush (Ceanothus spp.). In the riparian zone, quaking aspen (Populus tremuloides) can be common as well as mountain azalea (Rhododendron occidentale), red-osier dogwood

(Cornus sericea occidentalis), and willow (Salix spp.).

We selected tributaries of the Tuolumne River and Merced River based on burn characteristics including severity and time since last burn. Twelve tributaries (Figure 6) were selected and grouped in pairs by fire severity (i.e., each pair consisting of a low- severity and a high-severity burned reach; all fires were natural wilderness/non-

70 prescribed fires). Following Jackson and Sullivan (2009) and Malison and Baxter

(2010a), fire severity was determined at the site (i.e., stream reach) level based on the condition of the conifer canopy: low-severity – canopy intact, only the riparian understory was burned; high-severity – conifer canopy removed by fire over at least 75% of the reach). All stream reaches were burned between 1996 and 2011 (Table 1). In addition, a “reference” study reach was chosen on Chilnualna Creek, an area that has not experienced fire since at least 1930. We attempted to minimize variation in other variables including year since last burn, elevation, aspect, stream geomorphology, stream size, and dominant vegetation between paired reaches (Table 1), although accessibility also was a factor in determining study-reach selection. Stream reaches ranged in elevation from 1,574 to 2,154 m. All stream reaches exhibited high-gradient cascade or step-pool channel morphology (Montgomery and Buffington 1997) and represented tributaries of varying aspects in both river systems. Although study reaches comprising a few pairs were spatially distant from one another (e.g., 5a, b; Figure 6), we chose to focus on reducing variability in the aforementioned characteristics.

Each reach was established as approximately 10X bankfull width in length following Cianfrani et al. (2009). We identified comparable and representative stream reach types based on valley segment type, substrate type, reach gradient, and channel bankfull width and depth, and floodprone width (Montgomery and Buffington 1997,

Davis et al. 2001, Cianfrani et al. 2009). At each study reach, we surveyed channel geomorphology and composition of riparian vegetation.

Field Sampling

Sampling took place once at each study reach across two field seasons conducted in 2011 and 2012. Chilnualna Creek was surveyed in 2014, which was added to the experimental design in order to incorporate an unburned treatment into the study, thereby allowing us to identify any major differences that might result from wildfire of any severity. At each study reach, we established three cross-channel transects (upstream, midstream, downstream) at representative riffle, run, step, and/or pool features. At each transect, we measured bankfull width and height, bankfull depth, and floodprone width

(Cianfrani et al. 2009). Sediment size distributions were conducted at each transect (100 pebbles at each of the three transects) based on Wolman (1954) and embeddedness

(percent of fine sediment surrounding cobbles) was visually estimated for fifty cobbles per reach. Stream order was recorded based on Strahler (1957) and elevation, aspect, and slope were also recorded for each stream reach based on remotely-sensed data (10-m2 digital elevation models (DEM) acquired from a fire atlas maintained by YNP and created by the US Geological Survey).

Vegetation plots (rectangular plots 5 x 20 m with long axis parallel to the stream channel) were established within the riparian zone (one per bank) along the study reach.

The long axis of the plots was within 0.5 m from the water edge. Riparian zones in low- order mountain streams can be quite narrow, so tree plots frequently encompassed riparian and near-stream upland vegetation. Nested within each 5x20 tree plot, we established shrub and herb subplots (2 x 4 m and 1 x 1 m, respectively located in tree plot

72 corners). These sub-plots were completely contained within the riparian zone and dominated by riparian-obligate plant species. We visually estimated percent cover for all plant species (grasses, forbs, and woody plants). Owing to overlapping species (in space), percent cover using these methods can be over 100%. These methods follow widely-used sampling protocols (Muller-Dombois and Ellensburg 1974, Goldsmith et al. 1986,

Jackson and Sullivan 2009). We identified plant species using The Jepson Manual:

Higher Plants of California (Hickman 1993).

Numerical and statistical analysis

We calculated width-to-depth, entrenchment, and incision ratios from geomorphic field measurements (Rosgen 1996, Sullivan and Watzin 2008). Width to depth ratio

(bankfull width ÷ mean bankfull depth) is a key measure used to assess the available energy within the channel and the degree to which the channel may be widening.

Entrenchment ratio (floodprone width ÷ bankfull width, increasing values indicate less entrenchment) describes the degree to which a channel is inset in its valley. Incision ratios (low-bank height ÷ maximum bankfull depth) are supplemental measures of bed degradation and often signify initial channel bed downcutting.

We used paired t-tests to determine differences between percent cover in each of the tree, shrub, and herb layers and to detect differences among geomorphic parameters including D50 (mm); embeddedness (%); and width-to-depth, entrenchment; and incision ratios using JMP 10.0 (SAS Institute Inc., Cary, North Carolina, USA ). We tested

73 assumptions of normal distribution and homogeneity of variance and transformed data where transformations resulted in closer adherence to assumptions.

The proportion of each catchment that had been burned greater than two times since 1930 (i.e., fire frequency) and the proportion burned with moderate to high-severity during the most recent fire > 200 acres (i.e., fire extent) was determined for the area draining to each study reach (i.e., catchments). Each catchment was delineated using the watershed tool in ArcGIS 10.1 (Esri, Redlands, CA, USA) from 10-m2 DEM geospatial data acquired from the YNP fire atlas and created by the US Geological Survey. Fire frequency was determined from fire history data provided by YNP and available from the

Integrated Resoure Management Applications portal (available online at https://irma.nps.gov/). Burn severity was estimated at the catchment level using normalized burn ratio values (NBR) calculated from Landsat 7 Enhanced Thematic

Mapper Satellite Imagery following (Key and Benson 2006). Relative differences in normalized burn ratios (RdNBR) were calculated for each burned catchment. Breakpoints in RdNBR were determined for each pixel and assigned as unburned, low-severity, moderate-severity or high-severity. Percent catchment burned at each level of severity, and total percent catchment burned was determined from these estimates. We then used simple linear regression to explore the potential influence of these catchment-scale fire variables on stream geomorphic characteristics.

To identify possible differences in riparian plant community composition between burn categories, we used Non-metric Multidimensional Scaling (NMS) and Multi-

Response Permutation Procedure (MRPP) ordinations using PC-ORD 5 (McCune and

Mefford 1999). We converted field estimates of percent cover to an octave scale and used the midpoints of each octave class to minimize sampling errors while still preserving finescale differences (Muller-Dombois and Ellensburg 1974). We used the following octave classes: 1 (trace), 2 (0-1%), 3 (1-2%), 4 (2-5%), 5 (5-10%), 6 (10-25%), 7 (25-

50%), 8 (50-75%), 9 (75-95%), 10 (>95%). For riparian trees, shrubs, and herbs, we obtained an importance value from the square of the mean of relative frequency and relative cover estimates and used these values for ordination. We excluded rare species with an importance value less than 0.5% from the analysis following McCune et al.

(2002).

NMS ordination is suitable for analysis of plant community data (McCune et al.,

2002) as it does not make assumptions relative to the structure of the data and it preserves the distance between communities in ordination space better than other ordination techniques (McCune et al. 2002). We utilized Sorenson (Bray Curtis) distance for each

NMS, with a randomized-starting configuration and an initial ordination with a step down from six dimensions to determine the appropriate number of dimensions. The final starting configuration from the previous ordination was used to run the final ordination.

We conducted both initial and final ordinations with 50 runs of real data and a Monte-

Carlo test with 250 runs of randomized data. We obtained Pearson’s r rank coefficients for riparian vegetation and physical parameters including time since fire, elevation, stream geomorphic measures.

We used multivariate non-parametric procedure (MRPP) to test for differences in plant community composition and structure among burn types following NMS ordination

75 of each community. Where testing for potential differences among pre-defined groups

(e.g., high-severity burn vs. low-severity burn) is warranted, MRPP is a broadly utilized in ecological applications (see Mielke, 1984; Meilke and Berry, 2001 as cited in McCune et al., 2002). MRPP provides an A-statistic and p-value based on 250 Monte-Carlo simulations. We used Sorenson (Bray-Curtis) distances for this procedure as well, and applied a Bonferroni correction to account for multiple comparisons (Miller, 1981). We used PC-ORD software for both NMS and MRPP analyses (McCune et al., 2002).

Phase 2: BACIP experiment (2014)

The Rim Fire burned extensively (> 250,000 acres) and with large patches of high-severity in the summer of 2013. The burning of two of our 2011-2012 study reaches presented an opportunity for a before-after comparison. Therefore, we also present a before-after analysis of the short-term effects of the Rim Fire on riparian vegetation and stream geomorphology.

Following a BACIP (paired before-after, control-impact) design (Stewart-Oaten et al. 1986, Downes et al. 2002), we returned to two stream reaches that were burned by the

Rim Fire as well as two control reaches (Figure 7). Middle Tuolumne Creek was previously categorized as high-severity burned by a fire occurring in 1996, and South

Tuolumne was previously categorized as low-severity burned by a fire in 2002. Middle

Tuolumne was again burned by high-severity fire during the Rim Fire and South

Tuolumne with low-severity fire, thus creating two separate treatments (i.e., high/high and low/low). For control reaches, we resampled at Frog Creek and Cascade Creek

(neither of which were inside the Rim Fire perimeter), which were similar in elevation and stream size to Middle Tuolumne and South Tuolumne, respectively. Frog Creek was previously classified as high-severity while Cascade Creek was previously classified as low-severity. At these four reaches, we collected (and calculated) a subset of data

(riparian vegetation cover and composition and stream geomorphology) following the same protocols as in 2011-2012. Plot corners and transects were not marked with permanent markers. Therefore, we used a combination of geographic coordinates and field notes to approximate prior sampling locations. We used paired t-tests to compare our focal measures before (2012) and after (2014) the Rim Fire for each of the four study reaches. We calculated means and standard deviations for t-tests using intra-reach subsamples as “replicates”, whereby we partitioned percent cover of each vegetation type by each sub-plot (i.e., four per reach for herbs and shrubs) and utilized the three cross- sections and pebble counts within each reach for geomorphic comparisons. We could not partition richness or tree cover measures, so we drew conclusions based on a visual inspection of before-after results. For all other analyses, significance was determined at α

= 0.05, with α = 0.10 considered as evidence of a trend (Bocharova et al. 2013, Rowse et al. 2014).

Results

Phase 1: Paired design (2011 and 2012)

Geomorphology

Paired t-tests revealed no significant differences or trends between pairs for any geomorphic measures (p > 0.10 for all) (Figure 8). Across study reaches, D50 was similar at high-severity and low-severity burned reaches [9.2 ± 7.5 mm (SD) and 10.0 ± 4.7 mm

(SD), respectively]; however, there was modest separation of embeddedness values between treatments: 9.1 ± 4.9% (SD) at high-severity reaches compared to 6.7 ± 3.9%

(SD) at low-severity reaches. D50 was 7.8 ± 0.7 mm (SD) and embeddedness was 8.0 ±

6.5% (SD) at Chilnualna Creek (reference reach). Width-to-depth, entrenchment, and incision ratios at high-severity burned reaches were 13.0 ± 9.7 (SD); 11.8 ± 7.0 (SD); and

1.3 ± 1.23 (SD), respectively compared to 11.2 ± 5.0 (SD); 10.0 ± 5.3 (SD); and 0.6 ± 0.2

(SD), respectively at low-severity burned reaches. Mean width-to-depth ratio at

Chilnualna Creek was 23.8 ± 16.3 (SD), mean entrenchment ratio was 12.8 ± 1.6 (SD), and mean incision ratio was 0.5 ± 0.2 (SD).

Fire frequency (i.e., the proportion of each catchment burned more than twice after 1930) was positively correlated with embeddedness (p < 0.01, Figure 9a), and negatively correlated with D50 (p < 0.05, Figure 9b) and entrenchment (p < 0.01, Figure

9c) and width-to-depth (p < 0.05, Figure 9d) ratios. Fire extent (i.e., the proportion of the catchment burned with moderate or high-severity) was not related to any geomorphic variables (p > 0.10, data not shown).

Riparian vegetation

We observed a total of 57 species, 30 of which were forbs and the remaining 27 trees or shrubs. The most common species in descending order were lodgepole pine (P. contorta murrayana), willow (Salix spp.), red-osier dogwood (C. sericea occidentalis), cow parsnip (Heracleum lanatum), and red fir (A. concolor). Species richness was not significantly different between low- and high-severity burned study reaches, although the range was greater across low-severity burned reaches [14.2 ± 4.8 (SD) and 12.3 ± 2.3

(SD), respectively (t = 0.77, df = 5, p = 0.472)]. Buena Vista (high-severity, 9) and

Cascade (low-severity, 9) supported the lowest number of species. The greatest number of species was detected at Coyote (low-severity, 23) followed by Middle Tuolumne

(high-severity, 15) and South Tuolumne (low-severity, 14). At the reference reach on

Chilnualna Creek, we detected three tree species including red fir (A. concolor), lodgepole pine, and sugar pine (P. lambertiana); five shrubs including Spiraea desiflora, thimbleberry (Rubus parvifloris), willow, mountain azalea (R. occidentale), and red-osier dogwood; and one species of herbaceous plant (not identified at any of the other locations; Sierra shooting star (Dodecatheon jeffreyi)] in addition to several more common species [i.e., alum-root (Heuchera micrantha), western hawkweed (Hieracium albiflorum), and California vetch (Vicia californica)]. Out of 27 riparian woody species and 30 riparian herbaceous species detected, 20 and 26, respectively, had an importance value greater than 0.5 and were used in NMS and MRPP analysis.

Paired t-tests revealed that percent cover of herbaceous (t = -0.33, df = 5, p =

0.757) and shrub (t = 1.26, df = 5, p = 0.262) species did not differ significantly between

79 paired stream reaches. Tree cover was significantly different between pairs (t = 3.02, df =

5, p = 0.029), but this is not surprising given that site selection was partly determined based on the characteristics of the conifer canopy. Across study reaches, herbaceous cover was 93.0 ± 53.0% (SD) for high-severity burned reaches and 81.8 ± 41.1% (SD) for low-severity burned reaches. Shrub cover was 45.5 ± 43.1% (SD) along high-severity burned reaches and 70.0 ± 29.4% (SD) along low-severity burned reaches, and percent cover in the tree canopy was lower on average along high-severity burned reaches [14.5 ±

14.3% (SD)] compared to low-severity burned reaches [83.3 ± 56.5% (SD)] (Figure 10).

At the reference reach, percent cover in herb, shrub, and tree layers were 25.6, 184.0, and

31.4% respectively.

Riparian woody vegetation

Based on MRPP comparisons, there was no difference in composition of woody riparian vegetation between high- and low-severity reaches (A = -0.03, p = 0.809). NMS ordination echoed these results. The three-axis NMS solution (stress = 5.4, p = 0.01) represented 93.5% of the total variation: 34.5% on axis 1, 20.8% on axis 2, and 37.2% on axis 3 (Figure 11a). Axis 1 represented the occurrence of lodgepole pine and incense cedar as well as riparian shrubs like mountain azalea, alder, California dogwood, and black cottonwood. Axis 2 was highly correlated with red fir, and axis three was negatively correlated with the occurrence of lodgepole pine and species of currant.

Elevation was positively correlated (Pearson’s Rank Coefficient) with axis 1 (r =

0.79) along with lodgepole pine (P. contorta murrayana) (r = 0.64). Incense cedar

(Libocedrus decurrens), trembling aspen, California dogwood (C. nuttallii), mountain azalea (R. occidentale), and alder (Alnus rhombifolia) were all negatively correlated with axis 1 (r = -0.68; -0.59; -0.57; -0.57; and -0.59 respectively). Stream gradient (i.e., slope) was positively correlated with axis 2 (r = 0.54) as well as red fir (A. concolor) (r = 0.86), and mountain azalea (r = 0.56). Lodgepole pine was negatively correlated with axis 3 (r =

-0.84) as well as species of currant (Ribes spp.) (r = -0.61), and red fir, red-osier dogwood (C. sericea occidentalis), and mountain azalea were all positively correlated (r

= 0.65; 0.66; and 0.59, respectively) (Table 2).

Riparian herbaceous vegetation

Similar to the tree and shrub communities, the three-axis NMS solution (stress =

15.2, p = 0.85) for herbaceous vegetation represented 65.3% of the total variation: 13.1% on axis 1, 12.9% on axis 2, and 39.3% on axis 3 (Figure 11b). MRPP comparisons revealed no significant difference in composition of herbaceous riparian vegetation between high- and low-severity burned reaches (A = -0.003, p = 0.512).

Unidentified grasses, California strawberry (Fragaria virginiana), and western hawkweed (Hieracium albiflorum) were all positively correlated (Pearson’s Rank

Coefficient) with axis 1 (r = 0.69; 0.62; and 0.55, respectively). Incision ratio was negatively correlated with axis 2 (r = -0.53) as well as cow parsnip (Heracleum lanatum)

(r = -0.59) and alum root (Heuchera micrantha) (r = -0.55). Species of rush (Juncus spp.) were positively correlated with axis 2 (r = 0.56). Elevation was positively correlated with axis 3 (r = 0.52) and stream gradient was negatively correlated (r = -0.57). Pearly

81 everlasting (Anaphalis margaritacea), applegate’s paintbrush (Casttilleja applegatei), and vetch (Vicia californica) were positively correlated with axis 3 (r = 0.69; 0.58; and

0.58 respectively), and bracken fern (Pteridium aquilinum) and tiger lily (Lilium parvum) were negatively correlated (r = -0.56; and -0.51 respectively) (Table 2).

Phase 2: BACIP experiment (2014)

At all four study reaches, geomorphic characteristics did not change significantly when measured in 2014 following the Rim Fire (Table 3). In the summer following the

Rim Fire (2014), we observed a downward shift in percent cover in the herb, shrub, and tree layers at Middle Tuolumne (before high, after high), South Tuolumne (before low, after low), and Frog Creek (before high, after control); Cascade Creek (before low, after control) was the only location where we observed an upward shift in percent cover (Table

3, Figure 12). Few of these changes were significant, although shrub cover at South

Tuolumne was significantly reduced (47.3%, t = -3.29, df = 3, p = 0.046) and herbaceous cover exhibited a declining trend at Frog Creek (t = -2.41, df = 3, p = 0.095) (Table 3,

Figure 12). Although tree cover was reduced by 21% at Middle Tuolumne following the

Rim Fire and 55% at South Tuolumne, we could not assess these differences statistically.

Trees were absent from the Frog Creek plot in 2011 and 2014, and Cascade Creek exhibited a marginal reduction in tree cover that is likely due to capturing slightly different cover data following re-establishment of study plots.

The number of species detected was reduced at three locations (Appendum A).

Frog Creek exhibited the largest reduction (8 species: 6 herbs and 2 shrubs), then Middle

Tuolumne (5 species, 1 tree and 4 shrubs), and South Tuolumne (4 species: 2 trees and 2 herbs). At Cascade Creek, two species of herbs previously present were not detected in

2014; however, one new species of shrub (Rubus parvifloris) and one new species of tree

(Pinus contorta murrayana) were detected (Appendum A).

Discussion

Our paired comparison of stream geomorphology and riparian vegetation revealed only minor differences between reaches burned with high- and low-severity wildfire in the last 3-15 years. In addition, in a before-and-after comparison, we observed no change in stream geomorphic characteristics in the first growing season following the Rim Fire.

In contrast, fire-frequency at the catchment scale was significantly correlated with embeddedness as well as entrenchment, incision, and width-to-depth ratios at the reach scale. This study took place at a time of worsening drought, with precipitation and snowpack decreasing substantially over the study period. Given the importance of frequency, timing, and magnitude of precipitation and streamflow for influencing the effects of wildfire on erosion and sedimentation (Swanson 1981, Arkle et al. 2010,

Shakesby 2011), fire-severity at the reach scale may be less important for determining physical and structural attributes of stream-riparian ecosystems than catchment-level interactions between fires and drought.

Stream geomorphology

We expected removal of aboveground vegetation and organic matter in the riparian zone and adjacent upland following high-severity fire to lead to increased overland flow, erosion, and sedimentation and subsequent changes in the balance between sediment and flow [i.e., following Lane’s conceptual relationship of the dynamic equilibrium between the driving and resisting forces for channel change (Lane 1955)].

However, we detected no significant effect of fire-severity at the reach scale in the mid- to-long term following wildfire as in our paired analysis or immediately following the

Rim Fire (i.e., BACIP analysis). The reference reach at Chilnualna Creek exhibited by far the greatest width-to-depth ratio [23.8 ± 16.3 (SD)]. Although the slope of this reach was not particularly steep (0.04), the segment of stream directly upstream of our reference reach was characterized by very steep cascade morphology. A large portion of our reach had been scoured to bedrock. Therefore bank erosion and channel widening leading to a high width-to-depth ratio at this location is likely the result of high-power flows entering the reach.

The degree to which overland flow increases and leads to subsequent erosion and stream sedimentation following wildfire is generally dependent on soil type, vegetation, topography, fire severity, and especially timing, frequency, duration, and magnitude of post-fire precipitation (Scott and Vanwyk 1990, Verkaik et al. 2013a). Our study took place in a period of worsening drought in the central Sierra Nevada range of California.

Precipitation for the southern Sierra was 160% of average (1956-2005) for the 2010-2011 water year (wy), 60% of average for the 2011-2012 wy, 65% of average for the 2012-

2013 wy, and 50% of average for the 2013-2014 wy (California Data Exchange Center

2014). Snowpack has also been steadily decreasing from roughly 300% of the annual average in the 2010-2011 wy to 16% in the 2013-2014 wy (California Data Exchange

Center 2014). Therefore, reach-scale differences in channel morphology and deposition of fine sediments between pairs and following the Rim Fire may have been dampened by reduced precipitation and spring runoff. For example, Arkle et al. (2010) found that streams draining catchments that had experienced a greater proportion of high-severity wildfire in the riparian zone exhibited greater interannual variability in sediment loads and that magnitude of peak streamflow was the source of this variation. Further, Verkaik et al. (2013b) found that benthic invertebrate community composition (which is generally highly influenced by post-fire stream geomorphology; Minshall 2003) was more dependent on drought conditions than the occurrence of fire and Rugenski and Minshall

(2014) found that anticipated post-fire reductions in benthic invertebrate biomass and stream-bed algal production were not realized during a period of reduced peak spring runoff. Although not significant, modest visual differences between treatment groups in embeddedness and entrenchment, incision, and width-to-depth ratios (which all were slightly greater at high-severity burned reaches compared to low-severity burned reaches) may provide initial evidence of future increases in erosion and sedimentation at these locations.

Temporal variability may also have played a role. For example, previous research has indicated that post-fire sediment transport usually decreases within 2-3 years post-fire in Mediterranean-type systems (Cerda and Doerr 2005, Mayor et al. 2007) and Davis et

85 al. (1989) found that the first storm following a fire event filled stream pools with fine sediments, but a subsequent storm flushed these sediments out leaving the system in near pre-fire condition. Sediment export from a burned stream in southern California returned to pre-fire levels within two years, but again reduced precipitation may have facilitated this result (Coombs and Melack 2012), and Rehn (2010) found that, although high levels of sediments persisted in burned streams in southern California two years after fire, by the fourth spring following wildfire most of the sediment was transported out of the local systems. We measured stream geomorphology three to 15 years following wildfire, so some of the effects of fire-severity may have been ameliorated over time. In addition, our study was limited to small headwater streams, and the effects of wildfire over time might be expected to play out differently in streams of different sizes and catchment positions.

For instance, channel adjustments shifted gradually from upstream to downstream over the first five years following the 1988 Yellowstone fires, possibility related to pulses of fine sediment working through the system (Minshall et al. 1997). Integrating the temporal effects of fire across entire catchments will be an important step in understanding fire as a disturbance agent in stream-riparian ecosystems.

Large wood can also mediate the effects of wildfire on hydrology and geomorphology of streams (Minshall et al. 1997, Gresswell 1999, Gurnell et al. 2002).

Inputs of large wood to low-order streams (i.e., 1-3) have been shown to force morphologies (e.g., step-pool) and contribute significantly to retention of fine sediments

(Gurnell et al. 2002, Ryan et al. 2011). Modest increases in large wood in high-severity burned reaches may indicate that these streams may diverge relative to geomorphic

86 characteristics over time as sediments are trapped and banks are stabilized by large wood, however in a in a companion study (Jackson and Sullivan In press) conducted at the same

Yosemite study reaches, we observed that the proportion of the nearshore zone occupied by large wood was not significantly greater at high-severity burned sites compared to low-severity burned sites.

We did not detect any significant associations between reach-scale geomorphic variables and the extent of moderate-to-high severity wildfire at the catchment scale, which contrasts results by Legleiter et al. (2003), who found that channels with a greater percentage of burned area in their catchments exhibited smaller width-to-depth ratios, greater stream power, and lower bank failure. Conversely, we did find significant relationships between fire frequency at the catchment scale and embeddedness, and width-to-depth and entrenchment ratios (positive relationships, note that low entrenchment ratios describe highly entrenched streams) and median sediment size

(negative relationship). Taken together, these patterns suggest that streams draining catchments that have experienced frequent historic wildfire may exhibit greater discharge, bank erosion, and decreased sediment supply leading to reorganization of channel morphology and bed composition and a potential critical channel degradation threshold (Bull 1979, Legleiter et al. 2003). In France, Lavabre et al. (1993) found that small-plot erosion was sensitive to short-duration rainfall whereas catchment level export was more dependent on longer-term precipitation patterns. Therefore, although reach- scale variability in geomorphology following wildfires in the Yosemite system may diminish over time, frequent wildfires at the catchment scale may continue to affect

87 hydrogeomorphology downstream years to decades later. The importance of fire frequency at the catchment scale indicates that management activities that alter the frequency of natural fire disturbances may have strong consequences for stream channel structure with implications for both aquatic habitat and biota.

Riparian vegetation

We are unaware of any other study in a Mediterranean-type ecosystem where percent cover of riparian vegetation in tree, shrub, and herb layers have been measured in the mid-to-long term following wildfire. We found no significant differences between high- and low-severity burned pairs in species richness or percent cover in the herb and shrub layer. Many riparian plants exhibit life-history traits that allow them to recover quickly following flooding and these life-history traits may serve a similar function following wildfire (Dwire and Kauffman 2003). Riparian plants in the North American

West may also be adapted to historic wildfire (Barro and Conard 1991, Wisheu et al.

2000). This may particularly be the case in Mediterranean-climate systems where riparian plant communities would have been exposed to frequent historic wildfire; dendrochronological evidence suggests riparian zones in the Sierra Nevada mountains burned with approximately the same or even greater fire frequency compared to adjacent uplands (Van de Water and North 2010).

We found significantly greater percent cover by trees in low-severity burned reaches but this is to be expected as removal of the conifer canopy determined severity classification. Despite exhibiting roughly the same species composition as other reaches,

88 the reference reach on Chilnualna Creek had the lowest percent cover of herbs of all the study reaches, and shrub cover was 4x the average at high-severity burned reaches and

2.5x the average at low-severity burned reaches. Although we expected shrub cover to be greater in riparian areas burned with low-severity compared to those burned with high- severity, we did not see significant differences between pairs, however the unburned reach in our study had far greater shrub cover than either.

We expected significant reductions in tree and shrub cover at both the stream reaches affected by the Rim Fire. Although tree and shrub cover were substantially reduced at South Tuolumne following the Rim Fire, there were no significant changes at

Middle Tuolumne even though it was burned with greater severity. This may be due in part to the pre-fire condition of riparian vegetation at these locations as tree cover was already low at Middle Tuolumne following a high-severity fire that occurred in 1996.

Conversely, shrub cover prior to the Rim Fire at Middle Tuolumne was roughly twice the average for all high-severity burned reaches in this study and regeneration of riparian vegetation was extremely rapid at Middle Tuolumne following the Rim Fire. We observed stump sprouting in the adjacent upland of up to 20 cm six months following the

Rim Fire (Figure 13). This is consistent with evidence from other studies conducted in

Mediterranean-type ecosystems where side-slope vegetation can recover rapidly through stump-sprouting and may be stimulated to germinate by heat, smoke, or charred wood

(Barro and Conard 1991, Wisheu et al. 2000) and surviving riparian woody vegetation re- sprout from surviving below ground tissue (Davis et al. 1989, Bêche et al. 2005).

Our NMS and MRPP analysis revealed no separation of riparian plant community composition by fire-severity. In contrast, a study conducted in central Idaho revealed that riparian plant community composition was significantly different between high-severity burned reaches and both low-severity burned and unburned reaches five year after the

Diamond Point fire (Jackson and Sullivan 2009). These differences were evident for riparian woody and herbaceous vegetation as well as adjacent upland vegetation. In the first growing season following an extensive high-severity fire in southern California,

Bendix and Cowell (2010b) described a dramatic shift from riparian plant communities dominated by Alnus rhombifolia prior to the fire, to Populus fremontii and Quercus agrifolia dominating post-fire and copious re-sprouting of Salix spp. and Q. dumosa. This result contrasts with work by Kobziar and McBride (2006) and Bêche et al. (2005), who indicated that low-severity and prescribed fire (respectively) have little effect on riparian plant community composition in the short-term following fire.

As with stream geomorphology, time since fire may have also played a role in our results. We measured community composition 3-15 years post-fire. Riparian vegetation across biomes is highly adapted to disturbance and recovers quickly following floods and fires (Dwire and Kauffman 2003, Rood et al. 2007). There is some evidence that recovery times for Mediterranean-type riparian forests are half that for temperate riparian forests

(Verkaik et al. 2013a), which may partly explain why we observed no difference in community composition between pairs. Bendix and Cowell (2010b) argued that composition of riparian vegetation post-fire may depend on disturbance history, life- history traits, and dispersal capability. At some of our study reaches, sufficient time may

90 have passed that even non-resistant species and poor dispersers could have recovered at high-severity burned reaches.

Jackson and Sullivan (2009) also found that shifts in herbaceous community composition were primarily related to the proliferation of the invasive grass Bromus tectorum in the upland and both Phalaris arundinacea and Bromus techtorum in the riparian zone, and we expected higher incidence of invasive plants in our high-severity burned plots, however we did not detect Bromus techtorum in any of our riparian plots.

Although we observed Phalaris arundinacea at three study reaches; its importance was low enough that it did not correlate significantly with any of the axis generated by NMS.

All these study locations were within designated wilderness areas: one of these study reaches was located near a wilderness campsite (Camp Creek) and the other next to a legacy road (Crane Creek), which likely serve as vectors for transport of exotic species.

The third location was Middle Tuolumne following the Rim Fire. Although we found little evidence for an interaction between fire-severity and occurrence of invasive species in this study, the proliferation of invasive grasses in riparian zones (and uplands) following wildfires can lead to increased frequency (D'Antonio and Vitousek 1992) and spread (Coffman et al. 2010) of wildfire and is, therefore, of considerable interest to managers.

Linkages between riparian vegetation and stream geomorphology

Post-fire floods and sedimentation events may remove or bury riparian vegetation and reset successional dynamics (Dwire and Kauffman 2003, Pettit and Naiman 2007b,

91 a). This interaction has rarely been studied in Mediterranean-type systems, but Bendix and Cowell (2010a) found that flooding following a high-severity wildfire undermined dead and damaged riparian trees and caused them to topple. Therefore, we expected that changes in channel geometry (i.e., increased width-to-depth ratio and decreased slope and incision ration due to bank erosion and deposition of sediments) would correlate with riparian plant community composition and structure. We detected a minor negative correlation between incision ratio and axis 2 of the herbaceous plant community NMS.

Species of rush (i.e., importance value calculated from relative cover and relative frequency) were positively correlated with this axis while alum-root and cow-parsnip were negatively correlated. However, this result may have more to do with floodplain connectivity, microclimate, or other aspects of habitat than wildfire as we saw no significant relationship between fire-severity, frequency, or extent with incision.

Conclusions

In this study, we expected to observe significant differences in riparian vegetation community composition and structure as a result of high-severity or recent wildfire, however the lack of significant differences in riparian vegetation community composition and structure between pairs and immediately following the Rim Fire refuted our predictions. Composition and structure of riparian vegetation may not be as affected by high-severity wildfire in Mediterranean climates due to a combination of factors including more heterogeneous wildfire effects in riparian zones, rapid regeneration, and adaptations that make species resistant or resilient to wildfire.

While we expected to see significant differences in stream geomorphology as a result of high-severity or recent wildfire, in our system, no significant differences were observed between reaches paired by fire-severity. Fire-related sediment pulses are related to a variety of factors, fire-severity being only one (Shakesby and Doerr 2006) and fire- frequency at the catchment scale was correlated with sediments and channel geometry while fire-severity at the catchment scale was not. Sediment delivery generally decreases with time since fire (Shakesby and Doerr 2006, Rehn 2010, Coombs and Melack 2012) and timing, magnitude, and duration of post-fire floods heavily influence sediment delivery and channel alteration (Gresswell 1999, Arkle et al. 2010, Verkaik et al. 2013a).

Therefore, fire-related sediment pulses result as an interaction between wildfire and floods and these interactions take place at broad spatial and temporal scales. This may be part of the reason why we detected no significant changes to stream geomorphology following the Rim Fire – our measurements took place following an unusually dry winter and we focused our measurements at the reach scale, as do most studies that relate wildfire effects to stream-riparian ecosystems. Research that integrates hydrogeomorphological responses to fire over broader temporal and spatial scales will likely be more beneficial for understanding the complex interactions between wildfire, climate, and stream-riparian ecosystem structure.

The influence of fire-severity on stream-riparian ecosystems observed in other studies is diverse and far-reaching. Stream geomorphology and riparian vegetation structure and composition have myriad implications for stream-riparian ecosystem function, biodiversity, and conservation (Naiman et al. 2005). Post-fire scouring flows

93 and changes to bed composition can result temporarily in decreased substrate available for benthic algae, and shifts in benthic invertebrate community composition toward habitat-generalist r-selected species (Gresswell 1999, Minshall et al. 2001). Sediment delivery and debris flows following severe wildfires can create spawning habitat for certain fish and their prey (Rosenberger et al. 2011, Sestrich et al. 2011), however, deposition of fine sediments can result in unsuitable spawning habitat for certain salmonids (Sestrich et al. 2011) and alter the heat budgets of streams (Royer and

Minshall 1997, Dunham et al. 2007) leading to higher water temperatures which can also effect fish distribution (Beakes et al. 2014). Removal of riparian vegetation by high- severity wildfire can also increase the amount of solar radiation entering streams and increase water temperatures. In addition, reductions in riparian vegetation can interrupt subsidies of leaf litter to in-stream consumers (Jackson et al. 2012) resulting in a decrease in the relative proportion of benthic macroinvertebrates belonging to shredder functional feeding groups (Mihuc and Minshall 1995). However increased solar radiation entering streams can also trigger increased in-stream primary and secondary productivity leading to increased abundance of benthic macroinvertebrates which are important sources of food for fish, and increased emergence of adult aquatic insects which are disproportionately relied upon as a food source for riparian birds, bats, and spiders

(Baxter et al. 2005, Malison and Baxter 2010b). Further, opening of the riparian canopy by fire can increase the foraging ability of aerial insectivorous bats (Buchalski et al.

2013). However, we observed no significant shifts in riparian vegetation or geomorphology in response to wildfire. Therefore, additional research that integrates the

94 influence of fire severity, extent, and frequency over broader scales of time and space and measures interaction effects with other sources of disturbance (i.e., flood and drought) are needed to determine the pattern and process resulting in seemingly unpredictable and site-specific results and, therefore, better inform fire management for ecosystem conservation.

Acknowledgements

Funding was provided by NSF DEB-1401480 awarded to SMPS and BKJ, Bureau of

Land Management (14-3-01-37) award to SMPS, and The Ohio State University. In

Yosemite National Park we received help from Dr. G. Smith and K. Van Wagtendonk.

We appreciate field assistance received from D. Groff, M. Hickson, and M. Ledford.

Literature Cited

Andreu, V., A. C. Imeson, and J. L. Rubio. 2001. Temporal changes in soil aggregates and water erosion after a wildfire in a Mediterranean pine forest. Catena 44:69-84. Arkle, R. S., D. S. Pilliod, and K. Strickler. 2010. Fire, flow and dynamic equilibrium in stream macroinvertebrate communities. Freshwater Biology 55:299-314. Barro, S. C. and S. G. Conard. 1991. Fire effects on California chaparral systems - an overview. Environment International 17:135-149. Baxter, C. V., K. D. Fausch, and W. C. Saunders. 2005. Tangled webs: reciprocal flows of invertebrate prey link stream and riparian zones. Freshwater Biology 50:201- 220. Beakes, M. P., J. W. Moore, S. A. Hayes, and S. M. Sogard. 2014. Wildfire and the effects of shifting stream temperature on salmonids. Ecosphere 5. Bêche, L. A., S. L. Stephens, and V. H. Resh. 2005. Effects of prescribed fire on a Sierra Nevada (California, USA) stream and its riparian zone. Forest Ecology and Management 218:37-59. Benda, L., D. Miller, P. Bigelow, and K. Andras. 2003. Effects of post-wildfire erosion on channel environments, Boise River, Idaho. Forest Ecology and Management 178:105-119. Bendix, J. and C. M. Cowell. 2010a. Fire, floods and woody debris: interactions between biotic and geomorphic processes. Geomorphology 116:297-304. Bendix, J. and C. M. Cowell. 2010b. Impacts of wildfire on the composition and structure of riparian forests in southern California. Ecosystems 13:99-107. Bertoldi, W., N. A. Drake, and A. M. Gurnell. 2011. Interactions between river flows and colonizing vegetation on a braided river: Exploring spatial and temporal dynamics in vegetation cover using satellite data. Earth Surface Processes and Landforms 36:1474-1486. Bisson, P. A., B. E. Rieman, C. Luce, P. F. Hessburg, D. C. Lee, J. L. Kershner, G. H. Reeves, and R. E. Gresswell. 2003. Fire and aquatic ecosystems of the western USA: current knowledge and key questions. Forest Ecology and Management 178:213-229. Buchalski, M. R., J. B. Fontaine, P. A. Heady, III, J. P. Hayes, and W. F. Frick. 2013. Bat response to differing fire severity in mixed-conifer forest California, USA. PLoS ONE 8. Bull, W. B. 1979. Thresholds of critical power in streams. Geological Society of America Bulletin 90:453-464.

Cerda, A. and S. H. Doerr. 2005. Influence of vegetation recovery on soil hydrology and erodibility following fire: an 11-year investigation. International Journal of Wildland Fire 14:423-437. Cianfrani, C. M., S. M. P. Sullivan, W. C. Hession, and M. C. Watzin. 2009. Mixed stream channel morphologies: implications for fish community diversity. Aquatic Conservation: Marine and Freshwater Ecosystems 19:147-156. Coffman, G. C., R. F. Ambrose, and P. W. Rundel. 2010. Wildfire promotes dominance of invasive giant reed (Arundo donax) in riparian ecosystems. Biological Invasions 12:2723-2734. Coombs, J. S. and J. M. Melack. 2012. The initial impacts of a wildfire on hydrology and suspended sediment and nutrient export in California chaparral watersheds. Hydrological Processes. D'Antonio, C. M. and P. M. Vitousek. 1992. Biological invasions by exotic grasses, the grass/fire cycle, and global climate change. Annual Review of Ecology and Sytematics 23:63-89. Davis, F. W., E. A. Keller, A. Parikh, and J. Florsheim. 1989. Recovery of the chaparral riparian zone after wildfire. Pages 194-203 in D. L. Abell, editor. Proceedings of the California Riparian Systems Conference: Protection, Management, and Restoration for the 1990s. USDA General Technical Report Pacific Southwest, Davis, CA. Davis, J. C., G. W. Minshall, C. T. Robinson, and P. Landres. 2001. Monitoring wilderness stream ecosystems. USDA Forest Service, Ogden, UT. DeBano, L. F. 2000. The role of fire and soil heating on water repellency in wildland environments: a review. Journal of Hydrology 231:195-206. DeFlorio, M. J., D. W. Pierce, D. R. Cayan, and A. J. Miller. 2013. Western US extreme precipitation events and their relation to ENSO and PDO in CCSM4. Journal of Climate 26:4231-4243. Doerr, S. H., A. J. D. Ferreira, R. P. D. Walsh, R. A. Shakesby, G. Leighton-Boyce, and C. O. A. Coelho. 2003. Soil water repellency as a potential parameter in rainfall- runoff modelling: experimental evidence at point to catchment scales from Portugal. Hydrological Processes 17:363-377. Downes, B. J., L. A. Marmuta, P. G. Fairweather, D. P. Faith, M. J. Keough, P. S. Lake, B. D. Mapstone, and G. P. Quinn. 2002. Monitoring ecological impacts: concepts and practice in flowing waters. Cambridge University Press, Cambridge, UK. Dunham, J. B., A. E. Rosenberger, C. H. Luce, and B. E. Rieman. 2007. Influences of wildfire and channel reorganization on spatial and temporal variation in stream temperature and distribution of fish and amphibians. Ecosystems 10:335-346. Dwire, K. A. and J. B. Kauffman. 2003. Fire and riparian ecosystems in landscapes of the western USA. Forest Ecology and Management 178:61-74. 97

Goldsmith, F. B., C. M. Harrison, and A. J. Morton. 1986. Description and analysis of vegetation.in P. D. Moore and S. B. Chapman, editors. Methods in Plant Ecology. Wiley & Sons, New York. Gregory, S. V., F. J. Swanson, W. A. McKee, and K. W. Cummins. 1991. An ecosystem perspective of riparian zones; focus on links between land and water. Bioscience 41:540-552. Gresswell, R. E. 1999. Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193-221. Gurnell, A. M., H. Piegay, F. J. Swanson, and S. V. Gregory. 2002. Large wood and fluvial processes. Freshwater Biology 47:601-619. Halsey, R. W. 2008. Fire, chaparral, and survival in southern California. Sunbelt Publications, San Diego. Hickman, J. C., editor. 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. Jackson, B. K. and S. M. P. Sullivan. 2009. Influence of fire severity on riparian vegetation heterogeneity in an Idaho, U.S.A. wilderness. Forest Ecology and Management 259:24-32. Jackson, B. K. and S. M. P. Sullivan. In press. Responses of riparian tetragnathid spiders to wildfire in forested ecosystems of the California Mediterranean climate region, USA Freshwater Science. Jackson, B. K., S. M. P. Sullivan, and R. L. Malison. 2012. Wildfire severity mediates fluxes of plant material and terrestrial invertebrates to mountain streams. Forest Ecology and Management 278:27-34. Johansson, M. E. and C. Nilsson. 2002. Responses of riparian plants to flooding in free flowing and regulated rivers: an experimental study. Journal of Applied Ecology 39:917-986. Key, C.H. and N.C. Benson. 2006. Landscape assessment: ground measure of severity, the Composite Burn Index; and remote sensing of severity, the Normalized Burn Ratio. In D.C. Lutes; R.E. Keane; J.F. Caratti; C.H. Key; N.C. Benson; S. Sutherland; and L.J. Gangi. 2006. FIREMON: Fire effects monitoring and inventory system. USDA Forest Service, Rocky Mountain Research Station, Ogden, UT. Gen. Tech. Rep. RMRS-GTR-164-CD: LA 1-51. Kobziar, L. N. and J. R. McBride. 2006. Wildfire burn patterns and riparian vegetation response along two Sierra Nevada streams. Forest Ecology and Management 222:254-265. Lane, E. W. 1955. The importance of fluvial morphology in hydraulic engineering. Proceedings of the American Society of Civil Engineers 745:1-17.

Lavabre, J., D. S. Torres, and F. Cernesson. 1993. Changes in the hydrological response of a small Mediterranean basin a year after a wildfire. Journal of Hydrology 142:273-299. Legleiter, C. J., R. L. Lawrence, M. A. Fonstad, W. A. Marcus, and R. Aspinall. 2003. Fluvial response a decade after wildfire in the northern Yellowstone ecosystem: a spatially explicit analysis. Geomorphology 54:119-136. Malison, R. L. and C. V. Baxter. 2010a. Effects of wildfire of varying severity on benthic stream insect assemblages and emergence. Journal of the North American Benthological Society 29: 1324-1338. Malison, R. L. and C. V. Baxter. 2010b. The fire pulse: wildfire stimulates flux of aquatic prey to terrestrial habitats driving increases in riparian consumers. Canadian Journal of Fisheries and Aquatic Sciences 67:570-579. Mayor, A. G., S. Bautista, J. Llovet, and J. Bellot. 2007. Post-fire hydrological and erosional responses of a Mediterranean landscape: Seven years of catchment-scale dynamics. Catena 71:68-75. McCune, B., J. B. Grace, and D. L. Urban. 2002. Analysis of ecological communities. MjM Software Design, Gleneden Beach, OR. McCune, B., Mefford, M.J., 1999. PC-ORD: Multivariate Analysis of Ecological Data. MjM Software Design, Gleneden Beach, OR. McMichael, C. E., A. S. Hope, D. A. Roberts, and M. R. Anaya. 2004. Post-fire recovery of leaf area index in California chaparral: a remote sensing-chronosequence approach. International Journal of Remote Sensing 25:4743-4760. Mihuc, T. B. and G. W. Minshall. 1995. Trophic generalists vs trophic specialists - implications for food-web dynamics in post-fire streams. Ecology 76:2361-2372. Miller, R.G., 1981. Simultaneous Statistical Inference. Springer Verlag, pp. 6-8. Minshall, G. F., J. T. Brock, D. A. Andrews, and C. T. Robinson. 2001. Water quality, substratum and biotic responses of five central Idaho (USA) streams during the first year following the Mortar Creek fire. International Journal of Wildland Fire 10:185-199. Minshall, G. W. 2003. Responses of stream benthic macroinvertebrates to fire. Forest Ecology and Management 178:155-161. Minshall, G. W., J. T. Brock, and J. D. Varley. 1989. Widlfire and Yellowstone's stream ecosystems. Bioscience 39:707-718. Minshall, G. W., C. T. Robinson, and D. E. Lawrence. 1997. Postfire responses of lotic ecosystems in Yellowstone National Park, USA. Canadian Journal of Fisheries and Aquatic Sciences 54:2509-2525. Montgomery, D. R. and J. M. Buffington. 1997. Channel-reach morphology in mountain drainage basins. GSA Bulletin 109:596-611. 99

Muller-Dombois, D. and H. Ellensburg. 1974. Aims and methods of vegetation ecology. Wiley & Sons, New York. Naiman, R. J., H. Decamps, and M. E. McClain. 2005. Riparia: Ecology, Conservation, and Management of Streamside Communities. Elsevier Academic Press, Burlington, MA. Pausas, J. G. and S. Paula. 2012. Fuel shapes the fire-climate relationship: evidence from Mediterranean ecosystems. Global Ecology and Biogeography 21:1074-1082. Pettit, N. E. and R. J. Naiman. 2007a. Fire in the riparian zone: characteristics and ecological consequences. Ecosystems 10:673-687. Pettit, N. E. and R. J. Naiman. 2007b. Postfire response of flood-regenerating riparian vetation in a semi-arid landscape. Ecology 88:2094-2104. Poff, N. L. 1992. Why disturbances can be predictable: a prespective on the definition of disturbance in streams. Journal of the North American Benthological Society 11:86-92. Polvi, L. E. 2011. Geomorphic and process domain controls on riparian zones in the Colorado Front Range. Geomorphology 125:504. Rehn, A. C. 2010. The effects of wildfire on benthic macroinvertebrates in southern California. San Diego, CA. Rice, S., M. Stoffel, J. M. Turowski, and A. Wolf. 2012. Disturbance regimes at the interface of geomorphology and ecology. Earth Surface Processes and Landforms 37:1678-1682. Rood, S. B., L. A. Goater, J. M. Mahoney, C. M. Pearce, and D. G. Smith. 2007. Floods, fire, and ice: disturbance ecology of riparian cottonwoods. Canadian Journal of Botany-Revue Canadienne De Botanique 85:1019-1032. Rosenberger, A. E., J. B. Dunham, J. M. Buffington, and M. S. Wipfli. 2011. Persistent effects of wildfire and debris flows on the invertebrate prey base of Rainbow Trout in Idaho streams. Northwest Science 85:55-63. Rosgen, D.L., 1996. Applied River Morphology. Wildland Hydrology, Pagosa Springs, CO. Rowse, L. M., A. D. Rodewald, and S. M. P. Sullivan. 2014. Pathways and consequences of contaminant flux to Acadian flycatchers (Empidonax virescens) in urbanizing landscapes of Ohio, USA. Science of the Total Environment 485:461-467. Royer, T. V. and G. W. Minshall. 1997. Temperature patterns in small streams following wildfire. Archiv Fur Hydrobiologie 140:237-242. Rugenski, A. T. and G. W. Minshall. 2014. Climate-moderated responses to wildfire by macroinvertebrates and basal food resources in montane wilderness streams. Ecosphere 5.

100

Ryan, S. E., K. A. Dwire, and M. K. Dixon. 2011. Impacts of wildfire on runoff and sediment loads at Little Granite Creek, western Wyoming. Geomorphology 129:113-130. Sala, M., M. Soler, and M. Pradas. 1994. Temporal and spatial variations in runoff and erosion in burnt soils. Pages 1123-1134 the Second International Conference on Forest Fire Research, Coimbra, Portugal. Scott, D. F. 1993. The hydrological effects of fire in South-African mountain catchments. Journal of Hydrology 150:409-432. Scott, D. F. and D. B. Vanwyk. 1990. The effects of wildfire on soil wettability and hydrological behavior of an afforested catchment. Journal of Hydrology 121:239- 256. Sestrich, C. M., T. E. McMahon, and M. K. Young. 2011. Influence of fire on native and nonnative salmonid populations and habitat in a western montana basin. Transactions of the American Fisheries Society 140:136-146. Shakesby, R. A. 2011. Post-wildfire soil erosion in the Mediterranean: review and future research directions. Earth-Science Reviews 105:71-100. Shakesby, R. A. and S. H. Doerr. 2006. Wildfire as a hydrological and geomorphological agent. Earth-Science Reviews 74:269-307. Shaw, J. R. 2008. Linkages among watersheds, stream reaches, and riparian vegetation in dryland ephemeral stream networks. Journal of Hydrology 350:68. Stanford, J. A., M. S. Lorang, and F. R. Hauer. 2005. The shifting habitat mosaic of river ecosystems. Verh. Internat. Verein. Limnol. 29:123-136. Stanley, E. H., S. M. Powers, and N. R. Lottig. 2010. The evolving legacy of disturbance in stream ecology: concepts, contributions, and coming challenges. Journal of the North American Benthological Society 29:67-93. Stewart-Oaten, A., W. W. Murdoch, and K. R. Parker. 1986. Environmental-impact assessment - psuedoreplication in time. Ecology 67:929-940. Sullivan, S.M.P., Watzin, M.C., 2008. Relating stream physical habitat condition and concordance of biotic productivity across multiple taxa. Canadian Journal of Fisheries and Aquatic Sciences 65, 2667–2677. Swanson, F. J. 1981. Fire and geomorphic processes. U.S. Forest Service General Technical Report WO-26, Washington, DC. Van de Water, K. and M. North. 2010. Fire history of coniferous riparian forests in the Sierra Nevada. Forest Ecology and Management 260:384-395. Verkaik, I., M. Rieradevall, S. D. Cooper, J. M. Melack, T. L. Dudley, and N. Prat. 2013a. Fire as a disturbance in mediterranean climate streams. Hydrobiologia 719:353-382.

101

Verkaik, I., M. Vila-Escale, M. Rieradevall, and N. Prat. 2013b. Seasonal drought plays a stronger role than wildfire in shaping macroinvertebrate communities of Mediterranean streams. International Review of Hydrobiology 98:271-283. Vila-Escale, M., T. Vegas-Vilarrubia, and N. Prat. 2007. Release of polycyclic aromatic compounds into a Mediterranean creek (Catalonia, NE Spain) after a forest fire. Water Research 41:2171-2179. Wisheu, I. C., M. L. Rosenzweig, L. Olsvig-Whittaker, and A. Shmida. 2000. What makes nutrient-poor mediterranean heathlands so rich in plant diversity? Evolutionary Ecology Research 2:935-955. Wolman, M. G. 1954. A method of sampling coarse river-bed material. Transactions of the American Geophysical Union 35:951-956. Wondzell, S. M. and J. G. King. 2003. Postfire erosional processes in the Pacific Northwest and Rocky Mountain regions. Forest Ecology and Management 178:75-87.

102

Table 1. Stream reaches in Yosemite National Park, California were paired by fire severity with an attempt to minimize variation in

time since fire, elevation, aspect, stream channel morphology, and dominant vegetation. Last year burned refers to the last time the

study reach burned; in some cases there was a more recent fire elsewhere in the catchment. Channel-reach morphology was

classified following Montgomery and Buffington (1997). Stream order was determined based on Strahler (1957).

Paired reaches Burn Year Elevation Aspect Order Channel-reach Dominant upland vegetation severity burned (m) morphology Buena Vista High 2001 2154 NNW 3 Step-pool Subalpine (Lodgepole Pine)

103 Mono Low 2004 2096 N 3 Pool-riffle Subalpine (Lodgepole Pine)

Frog High 2006 1909 S 3 Step-pool Lodgepole Pine and Red Fir Cascade Low 2007 1864 SW 3 Step-Pool Lodgepole Pine and Red Fir

Middle Tuolumne High 1996 1766 SW 3 Pool-riffle Yellow Pine Belt South Tuolumne Low 2002 1716 NW 3 Step-Pool Yellow Pine Belt

Tamarack High 2009 1931 S 2 Step-Pool Lodgepole Pine and Red Fir Grouse Low 2009 1574 NW 3 Step-Pool Yellow Pine Belt

Continued 103

Table 1 continued

Crane High 2009 1850 E 2 Step-Pool Lodgepole Pine and Red Fir Coyote Low 2007 1836 SE 2 Step-Pool Lodgepole Pine and Red Fir

Meadow High 2005 2127 N 1 Pool-riffle Subalpine (Lodgepole Pine) Camp Low 2005 2098 N 1 Pool-riffle Subalpine (Lodgepole Pine)

Chilnualna Reference unburned 2008 W 3 Pool-riffle Lodgepole Pine and Red Fir

104

Parameters Pearson's r

Riparian woody vegetation Axis 1 Axis 2 Axis 3 Physical parameters

Width-to-depth ratio -0.19 -0.34 0.21

Entrenchment -0.26 -0.44 0.31

Incision 0.24 -0.24 -0.39

Elevation (m) 0.79 0.08 -0.33

Gradient -0.26 0.54 0.44

Time since last fire (years) -0.18 0.40 0.17

Species (r > 0.5)

Abies concolor 0.86 0.65

Pinus contorta murrayana 0.64 -0.84

Libocedrus decurrens -0.68

Cornus sericea occidentalis 0.66

Ribes spp. -0.61

Rhododendron occidentale -0.57 0.56 0.59

Alnus rhombifolia -0.59

Populus tremuloides -0.59

Continued

105

Table 2 continued

Cornus nuttallii -0.57

Salix spp. 0.54 -0.56

Riparian herbaceous vegetation Axis 1 Axis 2 Axis 3 Physical parameters

Width-to-depth ratio 0.22 -0.34 0.41

Entrenchment 0.18 -0.17 0.52

Incision -0.36 -0.53 0.16

Elevation (m) -0.36 -0.46 0.12

Gradient -0.15 0.22 -0.57

Time since last fire (years) 0.13 -0.14 0.24

Species (r > 0.5)

Anaphalis margaritacea 0.69 grasses 0.69

Fragaria vriginiana 0.62

Heracleum lanatum -0.59

Casttilleja applegatei 0.58

Vicia californica 0.58

Pteridium aquilinum -0.56

Juncus spp. 0.56

Heuchera micrantha -0.55

Hieracium albiflorum 0.55

Lilium parvum -0.51

106

Δ t p Middle Tuolumne (before high, after high) Herbaceous cover (%) -20.9 -0.69 0.537 Shrub cover (%) -51.8 -1.28 0.291 Tree cover (%) -21.0 Taxa richness -5

D50 (mm) 4.8 0.32 0.767 Embeddedness (%) -2.6 -0.76 0.489 Width-to-depth ratio 4.2 0.84 0.491 Entrenchment -1.4 -0.32 0.777 Incision -0.3 -2.54 0.127

Frog Creek (before high, after control) Herbaceous cover (%) -77.4 -2.41 0.095 Shrub cover (%) -18.3 -0.64 0.569 Tree cover (%) 2.0 Taxa richness -6

D50 (mm) 0.0 1.44 0.286 Embeddedness (%) 0.2 0.12 0.910 Width-to-depth ratio 6.2 0.45 0.700 Entrenchment -1.0 -0.22 0.846 Incision 0.1 0.91 0.460

Continued

107

Table 3 continued

Herbaceous cover (%) -62.3 -1.04 0.376 Shrub cover (%) -47.3 -3.29 0.046 Tree cover (%) -55.0 Taxa richness -4

D50 (mm) -0.1 0.88 0.472 Embeddedness (%) 0.2 0.11 0.920 Width-to-depth ratio -1.0 -0.32 0.776 Entrenchment -1.6 -1.16 0.366 Incision -0.2 -0.83 0.493

Cascade Creek (before low, after control) Herbaceous cover (%) 6.9 0.17 0.876 Shrub cover (%) 20.5 0.52 0.637 Tree cover (%) -10.0 Taxa richness -3

D50 (mm) 4.7 1.20 0.296 Embeddedness (%) 2.5 1.38 0.241 Width-to-depth ratio 5.2 1.66 0.239 Entrenchment 7.8 3.55 0.071 Incision 0.6 2.39 0.139

108

Table 4. Appendum A: Common herbaceous and woody species and their occurrence at each study reach. Site codes are as follows:

BV - Buena Vista, FR - Frog, MT - Middle Tuolumne, TA - Tamarack, CR - Crane, ME - Meadow, MO - Mono, CS - Cascade, ST

- South Tuolumne, GR - Grouse, CY - Coyote, CA - Camp, CH - Chilnualna. Chilnualna Creek is classified as a reference stream.

It was sampled in 2014 following the Rim Fire, but was unaffected by the fire, and has not burned significantly anywhere in the

catchment for > 80 years.

Species Reaches

109 High-severity Low-severity Post - Rim Fire

BV FR MT TA CR ME MO CS ST GR CY CA FR MT CS ST CH Trees and shrubs Abies concolor X X X X X X X X X Alnus rhombifolia X X Amelanchier alnifolia X Arctostaphylos manzanita X Betula oxidentalis X Castanopsis sempervirens X X Ceanothus cuneatus X Cornus nuttallii X X X Continued b 109

Table 4 continued

Cornus sericea occidentalis X X X X X X X X X X X X X Libocedrus decurrens X X X X X X Lonicera involucrata X X Pinus contorta murrayana X X X X X X X X X X X X Pinus lambertiana X X X X Pinus ponderosa X X X Populus tremuloides Populus balsamifera tricocarpa X Prunus emarginata X X X X X X 110 Pseudotsuga menziesii

Quercus lobata X X Rhododendron occidentale X X X X X X X X Ribes lacustre X Ribes spp. X X X X X X X X Rosa spp. X X Rubus parvifloris X X X X X Salix spp. X X X X X X X X X X X X X X Spiraea densiflora X X Symphoricarpos albus X

b Continued 110

Table 4 continued

Forbes Achillea millefolium X X X X X Anaphalis margaritacea X X X Casttilleja applegatei X Carex rostrata X Dodecatheon jeffreyi X Epilobium angustifolium X X X X X Equisetum arvense X X X X Erigeron foliosus 111 Erigeron Pergrinus X X X Erigeron breweri Fragaria californica X X Galium californicum X X X Geranium californicum X X Helenium bigeloni X X Heuchera micrantha X X X X X X X X X X X X Heracleum lanatum X X X X X Hieracium albiflorum X X X Juncus ensilifolius X X X X X X X X

b Continued 111

Table 4 continued

Lilium parvum X Lupinus breweri X Lupinus polyphyllus X X X X X X X Mentha arevensis X X Mertensia ciliate X Mimulus bicolor X Phacelia heterophylia X Phalaris arundinacea* X X X Pteridium aquilinum X X X X X 112 Solidago californica X X X X X X Thalictrum fendleri X X Vicia californica X X X X X X * Non-native species

112

Figure 5. Area burned by decade between 1931 and 2012 in Yosemite National

Park, California, USA.

113

Figure 6. Fire history by decade within Yosemite National Park, California.

Paired reaches are Tamarack (1a) and Grouse (1b); Meadow (2a) and Camp (2b);

Buena Vista (3a) and Mono (3b); Middle Tuolumne (4a) and South Tuolumne

(4b); Frog (5a) and Cascade (5b); and Crane (6a) and Coyote (6b). The reference study reach is (r) Chilnualna.

114

Figure 7. The 2013 Rim Fire perimeter and paired stream reaches sampled in

2011 and 2012. Middle Tuolumne and South Tuolumne were consumed by the

Rim Fire with high- and low-severity wildfire, respectively, and were used as impact study reaches in 2014. Frog and Cascade were unburned by the Rim Fire and used as high- and low-severity control reaches, respectively.

115

15 (mm)

50 D 5

Embeddedness Embeddedness (%) 0

Figure 8. Geomorphic measurements [width-to-depth, entrenchment, and incision ratios, D50 (mm); and embeddedness (%)] presented by study pair. Light-grey columns represent high-severity burned stream reaches, dark-grey columns represent low-severity burned reaches, and the black column represents the reference reach on Chilnualna Creek.

Continued 116

Figure 8 continued

20 depth ratio depth

- 15

to - 10

Width 5

1 Incision ratio Incision

0 Entrenchment ratio Entrenchment

117

Figure 9. Relationships between fire frequency (i.e., proportion of catchment burned more than twice after 1930) with (a) embeddedness (R2 = 0.55, F = 12.3, p

2 2 = 0.006); (b) D50 (R = 0.43, F = 7.67, p = 0.020); (c) entrenchment ratio (R =

0.56, F = 12.50, p = 0.005); and (d) width-to-depth ratio (R2 = 0.40, F = 6.52, p =

0.029).

118

(%) Embeddedness

(mm)

50 D

c Entrenchmentratio

depth ratio depth

- Width

Fire frequency

119

120

100

Herbaceous cover Percent cover (%) cover Percent 20 Shrub cover Tree cover

0 High severity Low severity

Figure 10. Mean (symbols) and standard error (bars) for % cover of herbs, shrubs, and trees between high-severity and low-severity burned study reaches.

Note that overlapping species (in space) can result in cover values that exceed

100%.

120

Shrub Shrub Shrub (a) 1.0 1.0

1.0 0.5 0.5

0.0 0.0

Axis 3 Axis 3

Axis 2 0.0

Axis 2 (20.8%) 2 Axis (37.2%) 3 Axis Axis 3 (37.2%) 3 Axis -0.5 -0.5

-1.0 -1.0 -1.0 -1.0 0.0 1.0 -1.0 0.0 1.0 -1.0 0.0 1.0 Axis 2 (20.8%) Axis 1 (34.5%) Axis 1 (34.5%) Axis 2 Axis 1 Axis 1

HerbOutput HerbOutput HerbOutput (b) 1.5 1.5

0.5

0.5 0.5

Axis 2 Axis 3 Axis 3 -0.5

-0.5 -0.5

Axis 3 (39.3%) 3 Axis (12.9%) 2 Axis Axis 3 (39.3%) 3 Axis

-1.5 -1.5 -1.5 -1.5 -0.5 0.5 -1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0 Axis 2 Axis 1 Axis 1 Axis 2 (12.9%) Axis 1 (13.1%) Axis 1 (13.1%)

Figure 11. 2-D representation of NMS ordination of (a) riparian woody vegetation; and (b) riparian herbaceous vegetation. Reaches classified as high- severity burned are indicated by crosses, low-severity burned by triangles, and the reference reach by a circle. The three-axis solution shown is a simplification of riparian species importance in community space. Visually there is little partitioning of riparian species importance indicating highly heterogeneous community composition across study locations irrespective of fire-severity classification.

121

Figure 12. Results of paired before-after control-impact (BACIP) analysis of riparian herb, shrub, and tree cover at two stream reaches burned by the Rim Fire and two control reaches. Means are represented by symbols, standard error by bars. The vertical dashed line represents the Rim Fire. * indicates significance at α

= 0.05. ** indicates a trend at α = 0.10. The treatments for each study reach are:

Middle Tuolumne – before high, after high; South Tuolumne – before low, after low; Frog – before high, after control; and Cascade – before low, after control.

Note that overlapping species (in space) can result in cover values that exceed

100%.

122

150 Middle Tuolumne South Tuolumne Frog 100 Cascade

50 Herb cover (%) cover Herb

0 *

150

100

** Shrub cover (%) cover Shrub

150

100

50 Tree cover (%) cover Tree

0 2012 2014

123

Figure 13. (a) Stump sprouting of the common upland plant Arctostaphylos manzanita in the side slope adjacent to Middle Tuolumne Creek six months following the Rim Fire. 120 x 180 mm notebook is shown for scale. Photo taken in March, 2014. (b) Proliferation of riparian vegetation at Middle Tuolumne

Creek 13 months following the Rim Fire. (Photo credit B. Jackson, October,

2014).

124

Chapter 3: Responses of riparian tetragnathid spiders to wildfire in forested ecosystems of the California Mediterranean climate region, USA

Breeanne K. Jackson; S. Mažeika P. Sullivan

Abstract: Mediterranean ecosystems of California are characterized by high interannual variability in precipitation and susceptibility to frequent high-intensity wildfires. These drivers of fire are likely to become more pronounced because of climate change, but their relative effects on linked aquatic-terrestrial components of Mediterranean ecosystems have received limited attention. We investigated the effects of wildfire on riparian spiders of the family Tetragnathidae, which are common shoreline consumers in stream ecosystems that can be highly reliant on aquatic food resources. From 2011-2012, we assessed stream geomorphology; density and community composition of aquatic benthic macroinvertebrates; and density, mercury (Hg) body loads, trophic position (TP), and reliance on aquatic energy (using naturally abundant carbon and nitrogen isotopes) of tetragnathid spiders in six high-severity and six low-severity paired stream reaches in

Yosemite National Park, USA. Following the 2013 Rim Fire, we resurveyed a subset of these variables in four reaches using a paired BACIP (before-after, control-impact) design. In addition, we explored how reach- and catchment-scale variability might affect spider density and trophic dynamics. Although differences in tetragnathid spider responses between paired reaches were mixed, model-selection results indicated that 125 variability in benthic invertebrate density, catchment-scale fire frequency, and precipitation were important drivers of tetragnathid spider density and trophic position.

The consistent signal of precipitation across multiple spider responses may indicate that climate variability could overwhelm the influence of fire on aquatic-terrestrial ecological linkages.

126

Introduction

Globally, wildfire activity is increasing in both scope and frequency (Westerling et al. 2006, Flannigan et al. 2009, Moritz et al. 2012), with climate change-induced shifts in terrestrial ecosystem structure and function implicated as a significant driver (Davis et al. 2013). A large portion of California is influenced by a Mediterranean-type climate including the west slope of the Sierra Nevada mountain range from 28-44°N (Grove and

Rackham, 2001). California Mediterranean-type climate is characterized by periods of extremely wet and dry conditions and high interannual variability in precipitation

(Bonada and Resh 2013). Streams in the Sierra Nevada generally have peak flows in the fall and again in the spring following snowmelt. The magnitude of flow each year is dependent on the El Niño/Southern Oscillation (ENSO), with high flows in El Niño years and low flows in La Niña years (Bonada and Resh 2013). In addition, uplands draining into Sierra Nevada streams are highly susceptible to drought and are projected to experience heightened frequency and intensity of wildfire (Lenihan et al. 2003). In fact,

Miller et al. (2009) found that the percentage of high-severity fire in Ponderosa pine and mixed conifer forests of the Sierra Nevada has increased two fold in the last two decades

(1984- 2010). However, despite a broad recognition of fire as a key source of disturbance in terrestrial ecosystems (Agee 1993, Carrion et al. 2003, Stephens et al. 2007), the role of fire in stream ecosystems of Mediterranean climates has garnered relatively less attention than in temperate climates (Verkaik et al. 2013a).

In particular, the role of fire relative to the ecological connections between land and water, which has received increasing attention in recent years (Spencer et al. 2003,

127

Malison and Baxter 2010, Jackson et al. 2012), has not been the focus of research to date in ecosystems influenced by Mediterranean climate. A growing body of literature suggests that streams and their adjacent riparian zones are tightly-linked through energy exchanges, and reciprocal transfers of energy are essential to maintain ecosystem functions. Transfers of energy between terrestrial and aquatic ecosystems represent important energetic pathways by which terrestrially-derived organic matter, nutrients, and biota fuel aquatic consumers (Covich et al. 1999, Power et al. 2004, Romero et al. 2005).

Flows of energy from aquatic to terrestrial systems also provide important nutritional subsidies to riparian and terrestrial food webs (Power and Rainey 2000, Henschel et al.

2001, Baxter et al. 2005). Aquatic insects that emerge from streams as adults (hereafter

“emergent insects”) represent an especially critical energy source for riparian consumers

(Murakami and Nakano 2002, Baxter et al. 2005). For instance, spiders of the family

Tetragnathidae (a widely distributed riparian consumer) can be highly reliant on aquatic insects as prey (Sanzone et al. 2003, Burdon and Harding 2008). Sanzone et al. (2003) found that riparian orb-weaving spiders (Araneidae and Tetragnathidae) obtained 100% of their carbon (C) and 39% of their nitrogen (N) from in-stream sources. Thus, in this study, we use riparian tetragnathid spiders as model organism to assess the impacts of wildfire on cross-boundary ecological linkages.

Background

The characteristic upslope to downslope (e.g., Hynes 1975) and upstream to downstream (e.g., Vannote et al. 1980) connectivity of stream-riparian ecosystems makes

128 it probable that the extent, severity, and frequency of fire at both reach and catchment scales (sensu Frissell et al. 1986) have implications for cross-boundary ecological linkages (Polis et al. 1997, Gresswell 1999), but studies relating wildfire effects to stream-riparian ecosystems typically consider fire severity only at the reach scale

(although see Frissell et al. 1986, Arkle et al. 2010). This multi-scale influence of wildfire may be pronounced in systems characterized by strong Mediterranean dry-wet seasonality, high interannual variability in precipitation and runoff, and floods that frequently occur in fall and winter (often immediately following wildfire) rather than following spring snowmelt as in temperate ecosystems (Verkaik et al. 2013a). The strong interactions between wildfire and hydrology in governing stream ecosystem processes

(Arkle et al. 2010, Verkaik et al. 2013b, Rugenski and Minshall 2014) also implicate upstream wildfire characteristics as a likely driver of stream-riparian connectivity in

Mediterranean climate-mediated ecosystems.

High-severity fire (e.g., where the riparian conifer canopy is removed by fire) affects riparian plant community structure, composition, and distribution (Davis et al.

1989, Jackson and Sullivan 2009), and other aspects of stream-riparian systems such as large wood (Davis et al. 1989, Bendix and Cowell 2010, Vaz et al. 2011), directly modifying riparian spider habitat. Riparian plants are often highly adapted to fire disturbance (Dwire and Kauffman 2003, Jackson and Sullivan 2009) and in some cases can reestablish within a single growing season (Davis et al. 1989, Bêche et al. 2005). In addition, estimates of snag fall have ranged from 17-43% over 2-3 years in California shrublands (Davis et al. 1989, Bendix and Cowell 2010). Together, the recovery of

129 riparian vegetation and addition of large wood may result in greater habitat heterogeneity and web-building structure for riparian spiders. Conversely, low-severity fire (e.g., where the riparian canopy remains intact and only understory vegetation is consumed) may have little to no effect on riparian vegetation and inputs of large wood (and consequently tetragnathid spiders) over time (Jackson and Sullivan 2009, Arkle and Pilliod 2010).

Fire-severity might also be expected to influence energetic relationships within linked stream-riparian ecological networks. Trophic position (TP) is an integrative measure of food webs, as it reflects an organism’s relative feeding relationship within a community as well as the complexity of the web (Post 2002). For instance, maximum trophic position in a food chain, or food-chain length (FCL), describes the number of transfers of energy from the bottom to the top of a food web (Sabo et al. 2009) and is one of the primary indicators of ecosystem function and ultimately, stability. The dynamic constraint hypothesis (as cited in Pimm 2002) predicts that ecosystems affected by frequent or intense (i.e., exerting significant force) disturbance should have shorter FCL due to removal of predators, decreased biodiversity , and increased omnivory. Habitat heterogeneity has also been implicated as an environmental determinant of FCL (Persson et al. 1992). The influence of disturbance on FCL in stream ecosystems has typically been framed from a hydrological perspective [e.g., variability in discharge (Poff and

Ward 1990, Fausch et al. 2001, Sabo et al. 2010)] and to our knowledge terrestrial disturbance processes (e.g., wildfire) have not been considered within this context for streams. Although fish are commonly considered to be the top predators in aquatic systems and are most commonly used to estimate FCL (e.g., McHugh et al. 2010, Sabo et

130 al. 2010), in small tributary streams without piscivorous fish species, trophic position of tetragnathid spiders is likely on par with that of insectivorous fish and other aquatic vertebrate consumers [e.g., salamanders (Parker 1994)].

Both reach-scale and catchment-scale variability can affect the aquatic invertebrate communities (reviewed in Minshall 2003) on which tetragthanid spiders depend. Although changes in channel geometry following wildfire can be variable

(Verkaik et al. 2013a), fire-induced changes in stream hydrogeomorphology [i.e., greater stream discharge, reduced sediment supply, and streambed incision (Legleiter et al. 2003,

May and Gresswell 2003, Shakesby 2011)] can alter benthic macroinvertebrate larval populations with concomitant effects on adult emergence. In particular, loss of riparian leaf litter inputs (Jackson et al. 2012) and scouring flows (Koetsier et al. 2010, Vieira et al. 2011) can decrease shredder abundance and lead to shifts in the diet of individual taxa

(Mihuc and Minshall 1995, Spencer et al. 2003, Mihuc and Minshall 2005). In small headwater streams, higher densities of emergent insects have been associated with burned catchments (Mellon et al. 2008). Collectively, these patterns may result in greater reliance on aquatically-derived energy (i.e., nutritional subsidies originating from aquatic primary productivity) and a shift in spider TP.

Concentration of heavy metals [e.g., mercury (Hg)] in riparian consumers may serve as a complementary aquatic-terrestrial food-web tracer (Walters et al. 2008,

Walters et al. 2010, Alberts et al. 2013). Few if any aquatic ecosystems have escaped contamination, as even remote freshwater systems receive atmospherically-transported contaminants (Blais 2005). Contaminants in the tissues of aquatic insects are transported

131 to riparian and terrestrial consumers via multiple energetic pathways (Sullivan and

Rodewald 2012). For example, Walters et al. (2008) found that PCB concentrations in riparian spiders and herptiles were closely related to reliance on emergent aquatic prey.

Therefore, Hg concentration in riparian consumers can be indicative of reliance on emergent insect prey.

Approach

Within this linked stream-riparian context, we sought to understand how wildfire severity at the reach scale (for this study, 102 m), and wildfire frequency and extent at the catchment scale influence local stream-riparian food-web dynamics in a Mediterranean climate-mediated system. To do this, we evaluated the responses of Tetragnathidae to fire severity in six stream study reaches (one characterized as low severity and one as high severity within each pair) in Yosemite National Park (YNP) California, USA (2011-

2012) (Figure 14).

We used the naturally-abundant stable isotopes 13C and 15N to estimate realized

(sensu Post 2002) spider TP and reliance on aquatic energy pathways. In addition, we used the concentration of Hg in spider tissues as a complementary tracer of aquatically- derived energy sources. In high-severity study reaches, we predicted higher tetragnathid density and reliance on aquatically-derived energy and lower tetragnathid spider TP and

Hg body loading. We also assessed differences in geomorphology and aquatic food resources (benthic macroinvertebrates) between treatment groups.

132

To complement our categorical paired-reach design and to incorporate catchment- scale features, we also assessed the potential influences of a suite of quantitative, continuous variables at both the reach (i.e., geomorphology, riparian spider habitat, and density and community composition of benthic macroinvertebrates) and catchment scales

(i.e., fire extent, fire severity, catchment size, and precipitation) on tetragnathid spider responses. We predicted that: (1) rapid recovery of riparian vegetation and increased streamside habitat heterogeneity related to additions of small and large wood following high-severity fire would lead to increased web-building habitat structure for spiders and drive increases in spider density and increased reliance on aquatic energy ; (2) alternatively, changes in channel geometry and deposition of fine sediments associated with both the frequency and severity of wildfire would decouple nearshore habitat from the stream and reduce benthic macroinvertebrate density, respectively, leading to reduced

Tetragnathidae access to aquatic insect prey and decreased reliance on aquatically- derived energy ; (3) reductions in the diversity of benthic macroinvertebrate (used as a proxy for emergent insects, see “Methods”) assemblages following high-severity fire and shifts towards dominance by habitat generalists would result in reduced food-web complexity and lower tetragnathid TP; and (4) increases in fire frequency and extent would lead to decreased variability of TP through decreased habitat heterogeneity. The strength of evidence supporting each of these predictions was assessed using all 12 study sites in a model-selection approach (Burnham and Anderson 2004).

In the late summer of 2013, the Rim Fire, which burned extensively throughout the Stanislaus National Forest and Yosemite National Park (> 250,000 acres) and with

133 large patches of high-severity fire, burned two 2011-2012 study reaches (Figure 15). This presented an opportunity for a before-after comparison of the short-term effects of fire on stream-riparian ecosystem attributes. Thus, we also present evidence from a subset of our study reaches related to the effects of the Rim Fire on stream geomorphology, benthic macroinvertebrates, and tetragnathid spiders.

Methods

Phase 1: 2011-2012

Yosemite National Park has an excellent fire-history record and contains two large catchments located within an expansive wilderness largely devoid of confounding anthropogenic effects. We selected 12 tributaries of the Tuolumne River and Merced

River based on burn characteristics including severity and time since last burn (Figure

14). Within each stream, we selected a study reach as our sampling unit, defined here as approximately 10X bankfull width (Cianfrani et al. 2009). Study reaches were grouped in pairs by fire severity (i.e., each pair consisting of a low-severity and a high-severity burned segment; all fires were natural wilderness/non-prescribed fires). Fire severity was determined based on the appearance of the conifer canopy (i.e., low-severity: canopy intact, only the riparian understory was burned; high-severity: conifer canopy removed by fire over at least 75% of the study reach) (Jackson and Sullivan 2009, Malison and Baxter

2010a). All study reaches burned between 1996 and 2011 (Table 5). Although logistical constraints related to accessibility partly determined selection, we attempted to minimize variation in other variables including year since last burn, elevation, aspect, stream

134 geomorphology, channel width, drainage area, and dominant vegetation between paired reaches (Table 5). Although study reaches comprising a few pairs were spatially distant from one another (e.g., 5a and 5b; Figure 14), we chose to focus on reducing variability in the previously mentioned characteristics and examine the effects of regional climate by considering precipitation in our analysis.

In each study reach, we established three cross-channel transects (upstream, midstream, downstream). At each transect, we measured bankfull width and bankfull depth, and calculated width-to-depth ratio following Cianfrani et al. (2009). Median grain size (D50, mm) was estimated using Wolman pebble counts (1954) on 100 sediment clasts per transect (n = 300 per study reach). Percent of each study reach occupied by large wood (LW >10 cm diameter x > 1.0 m) and small wood (LW <10 cm diameter) was measured and counted. In addition, we recorded the percent of the nearshore zone (<1 m from water surface) of each study reach covered by understory vegetation.

Because of the wilderness setting of our study, we were unable to sample emergent insects directly with the use of emergence traps. Thus, we used benthic larval insects as a proxy, utilizing a rapid assessment adapted from the US EPA Rapid

Bioassessment Protocols (Barbour et al. 1999) to describe benthic macroinvertebrate community composition in the field. Macroinvertebrates were collected using a Surber sampler (500-µ mesh net) at 2-3 locations (depending on channel width) along each transect established for the geomorphic surveys (n = 6-9 samples per study reach).

Macroinvertebrates were identified to order using Voshell (2002) as a guide.

135

Tetragnathid spiders were surveyed along a 30-m transect running longitudinally along each bank. Transect locations were selected to be representative of the vegetation and hydrogeomorphic characteristics of each study reach at large based on the geomorphic surveys (as described above) and from vegetation surveys (Jackson and

Sullivan, In review). All tetragnathid spiders were counted for 60 minutes per bank at night (21:00-23:00) in July and August of 2011 and 2012 when spiders are at peak abundance in temperate regions (Williams et al. 1995, Meyer and Sullivan 2013). Four to eight spiders, one composite sample of epilithic algae collected from cobbles, and one composite sample of detritus were also collected from throughout the study reach for stable isotope analysis.

The area draining each study reach was delineated using the watershed tool in

ArcGIS 10.1 (Environmental System Research Institute, Redlands, California, USA). We then determined the proportion of each catchment that had been burned > twice since

1930 (i.e., fire frequency) and the proportion burned with moderate to high-severity during the most recent fire > 200 acres (i.e., fire extent). Burn severity was estimated at the catchment level using normalized burn ratio values (NBR) calculated from Landsat 7

Enhanced Thematic Mapper Satellite Imagery following (Key and Benson 2006).

Relative differences in normalized burn ratios (RdNBR) were calculated for each burned catchment. Breakpoints in RdNBR were determined for each pixel and assigned as unburned, low-severity, moderate-severity or high-severity. From these estimates, percent catchment burned at each level of severity, and total percent catchment burned was determined.

136

Precipitation data (cm) were gathered from the Western Regional Climate Center, using the closest Remote Automatic Weather Station to each study reach that was also situated at approximately the same elevation. Precipitation was determined based on average monthly precipitation for the water year prior to sampling [i.e., 01 October of the previous year to 30 September of the sampling year (Bêche and Resh 2007)].

Sample processing

In the laboratory, tetragnathid spiders were rinsed with distilled water and oven- dried at 60°C for 48 hours. Tissue from 3-5 individual spiders was then homogenized, pulverized using a mortar and pestle, and packed in tin capsules. Epilithic algae and terrestrial detritus were sorted from other materials (e.g., sediment and invertebrates), rinsed with distilled water, and then oven-dried. Subsamples of epilithic algae and terrestrial detritus were combined per study reach, respectively, and then homogenized into a fine powder using a Pica Blender Mill (Cianflone Scientific Instruments

Corporation, Pittsburgh, Pennsylvania, USA) or a mortar and pestle before packing composite samples per study reach for isotope analysis.

All samples were analyzed for 13C and 15N by continuous flow isotope-ratio mass spectrometry (EA-IRMS) at the Stable Isotope Core (Washington State University,

Pullman, Washington, USA). The results are reported in δ (‰) notation defined as:

δ13C or δ15N = [(Rsample/Rstandard)-1] * 100 where R is 13C/12C or 15N/14N, respectively. Typical analytical precision was 0.08‰ for δ15N and 0.19‰ for δ13C determination.

137

Estimating trophic position and reliance on aquatically-derived energy: After processing, we estimated spider trophic position using the two-source food web model from Post (2002); TP = λ + {δc – [δb1 / α + δb2 / (1-α)]} / Δn, where λ is the TP of the basal food sources (i.e., 1 for primary producers); δc is the δ15N signature of the consumer; δb1 and δb2 are the signatures of the two basal food sources; α is the proportion of N from basal food source 1; and Δn is the enrichment in δ15N per trophic level (i.e., 3.4 ‰; Post 2002). The proportion of N derived from basal source 1 (i.e., α) was estimated using a two-end member Bayesian isotopic mixing model solved with the

R software package SIAR [Stable Isotope Analysis in R; (Parnell and Jackson 2013)].

The SIAR package is equipped to handle variability in sources, consumers, and trophic fractionation factors (Parnell et al. 2010).

For all study reaches, the two end members were the basal food sources: epilithic algae and terrestrial detritus. The δ13C and δ15N data were used to estimate the contribution from each basal food source to the consumer. Trophic fractionation factors for tetragnathid spiders were estimated using the per trophic step fractionation in Post

(2002) (i.e., 3.4 ‰ ± 0.98 ‰ for δ15N and 0.39 ‰ ± 1.3 ‰ for δ13C) multiplied by the estimated number of trophic transfers between the consumer and basal resources

(estimated a priori as the difference between the consumer δ15N and mean basal resource

δ15N divided by 3.4 ‰; note however, in this study, adjusting for the number of trophic transfers actually had little effect on the mixing model results, as δ15N signatures of the two basal resources tended to be similar).

138

Tetragnathid spider tissue samples were analyzed for Hg concentrations [μg kg-1 dry weight] at the Diagnostic Center for Population and Animal Health, Toxicology

Section, Michigan State University. ICP-AES (Varian Vista, now part of Agilent, Santa

Clara, CA) and ICP-MS (Agilent 7500ce) instruments were calibrated with standards derived from NIST-traceable stock solutions for each element (GFS Chemicals, Inc.,

Cincinnati, Ohio). Quality was assured in each sequence run by analyzing lab reagent blanks, NIST (National Institute of Standards & Technology, Gaithersburg, MD)- traceable Multi-mix (Alfa Aesar Specpure, Ward Hill, Massachussetts) and digests of

NIST Standard Reference Materials (SRM), including Mussel Tissue 2976, Trace

Elements in Water 1643e and/or Montana II Soil 2711a, as appropriate. Mercury was analyzed by cold vapor atomic absorption spectrometry (CETAC CVAA, Omaha,

Nebraska) with similar calibration and quality control.

Statistical Analysis

We compared density, reliance on aquatically-derived energy, Hg body loading, and TP (as well as standard deviation of TP) of tetragnathid spiders between high- severity and low-severity study reaches using paired t-tests. We utilized both the mean and standard deviation of TP to ascertain differences in both magnitude and variability of spider TP between treatment groups. We also used paired t-tests to evaluate potential differences in benthic macroinvertebrate density and composition [i.e., % Ephemeroptera,

Plecoptera, Trichoptera (EPT)] and stream geomorphology (i.e., D50, width-to-depth ratio, and entrenchment ratio).

139

To evaluate the relative explanatory power of both quantitative reach- and catchment-level environmental variables on potential spider responses, we evaluated the effects of streamside habitat and geomorphology, benthic macroinvertebrate density and composition, catchment fire frequency and extent, catchment size, and precipitation on tetragnathid spider density, TP, reliance on aquatically-derived energy, and Hg body loading using a model-selection approach based on the Akaike Information Criterion adjusted for small sample size (AICc; Anderson and Burnham 2002, Burnham and

Anderson 2004). Competing models were selected per response variable based on the significance (p < 0.05) of the explanatory factors by linear regression, using the above descriptors. For comparison, the null model (i.e., intercept only) was also included in each set of competing models.

We first examined potential correlations among the predictor variables for each model set, and no variables with |r| ≥ 0.80 were included in the same model (e.g.,

Sullivan et al. 2007, Allen and Vaughn 2010). We ranked all candidate models according to their AICc values, and the most parsimonious model was the one with the lowest AICc value (Anderson and Burnham 2002, Burnham and Anderson 2004). We considered plausible models as those that were within 4 AIC units from the best model and calculated Akaike weights (wi) to determine the weight of evidence in favor of each model (Anderson and Burnham 2002, Burnham and Anderson 2004).

Subsequent linear regression was used to highlight key bivariate relationships between tetragnathid spider descriptors and environmental predictors and between tetragnathid TP and Hg body loading. Given our relatively limited sample size and high

140 variability in responses expected among study reaches, we used α = 0.05 to indicate statistical significance and α = 0.10 as a trend (e.g. Rowse et al. 2014). We performed all statistical analyses using JMP 10.0 (SAS Institute Inc., Cary, North Carolina, USA ).

Phase 2: BACIP experiment - 2014

Using a BACIP (paired before-after, control-impact) design (Stewart-Oaten et al.

1986, Downes et al. 2002), we compared two study reaches that were burned by the Rim

Fire with two study reaches that were not burned by the Rim Fire (i.e., control reaches).

Middle Tuolumne Creek was previously categorized as high-severity burned by a fire occurring in 1996, and South Tuolumne was previously categorized as low-severity burned by a fire in 2002. Middle Tuolumne was again burned by high-severity fire during the Rim Fire and South Tuolumne with low-severity fire, thus creating two separate treatments (i.e., high/high and low/low). For control reaches, we resampled at Frog Creek and Cascade Creek (neither of which were inside the Rim Fire perimeter), which were similar in elevation and stream size to Middle Tuolumne and South Tuolumne, respectively. Frog Creek was previously classified as high-severity while Cascade Creek was previously classified as low-severity.

We collected (and calculated) a subset of data (tetragnathid spider density, benthic invertebrate density, D50, % large and small wood, and % overhanging vegetation) following the same protocols as described above. We used paired t-tests to compare our focal measures before (in 2012) and after (in 2014) the Rim Fire for each of the four study reaches. We calculated means and standard deviations for t-test analysis using

141 intra-site subsamples as “replicates”, whereby we partitioned spider density, % overhanging vegetation, and % large and small wood by 5-m increments and by bank and calculated D50 for each of the three pebble counts per study reach. In addition, we utilized benthic invertebrate density within each taxonomic order to generate means and standard deviations for t-tests.

Results

Phase 1: 2011-2013

Average monthly precipitation for each water year ranged from 3.9 cm month-1

(Crane Flat, 2012) to 13.5 cm month-1 (Wawona, 2011) (Table 6; Figure 16). Most of the winter precipitation falls as snow; July and August were the driest months. Mean monthly precipitation at Crane Flat decreased throughout the study period from 8.1 cm month-1 in the 2010-2011 water year to 3.5 cm month-1 in the 2013-2014 water year.

Study reaches ranged from 1st to 3rd order and varied in size from 1.3 to 14.0 m wide at bankfull (Table 6). Mean depth varied from 0.9 to 0.3 m. The percent of small and large wood in high-severity burned study reaches was almost twice that of low- severity burned reaches (t = -2.0, df = 5, p = 0.103; Figure 17). Percent overhanging vegetation, width-to-depth ratios, and D50 did not differ between treatment groups

(Figure 17).

Benthic macroinvertebrate density was 1.5 times greater in high-severity burned reaches (t = -2.3, df = 5, p = 0.072; Figure 17). We identified benthic invertebrates from the orders Megaloptera, Plecoptera, Trichoptera, Ephemeroptera, Coleoptera, Diptera,

142

Trombidiformes (Hydracarina), Oligochaeta, and class Tubellaria. The majority of individuals belonged to the EPT orders (72% +/- 11% across all study reaches). This pattern was consistent across treatment groups (Figure 17).

Although tetragnathid spider density was numerically higher in low-severity burned study reaches compared with high-severity burned reaches, this difference was not significant (t = 1.5, df = 5, p = 0.195: Figure 17). Tetragnathid reliance on aquatically-derived energy, TP, and body loading of Hg were also not significantly different between treatment groups. Trophic position was a weak predictor of tetragnathid

Hg loading across all study reaches (R2 = 0.30; Figure 18), confirming that emergent insects contributed to tetragnathid spider diet.

Model-selection results suggested support for two models explaining tetragnathid spider density (Table 7). In the model receiving the most support (wi = 0.58), precipitation and benthic macroinvertebrate density explained 78% of the variation in tetragnathid spider density across study reaches. The second best model (wi = 0.23) – with precipitation as a predictor variable – explained 53% of the variation in tetragnathid spider density. For tetragnathid spider TP, we found support for models including both benthic macroinvertebrate density and fire frequency. A bivariate model with benthic macroinvertebrate density and fire frequency as predictor variables received the greatest support for explaining TP (wi = 0.50) and a univariate fire frequency model received the greatest support for explaining standard deviation of TP (wi = 0.40). For both tetragnathid reliance on aquatically-derived energy and body loading of Hg, the null models were most strongly supported, receiving 78% and 68% of support in their respective model sets

143

(Table 7). However, for both of these response variables, univariate models with precipitation also received support (wi = 0.19 for reliance on aquatically-derived energy and wi = 0.25 for Hg body loading) and warrant consideration as plausible models.

Phase 2: BACIP design - 2014

Two of the four study reaches sampled in the summer following the Rim Fire exhibited a reduction in tetragnathid density (p < 0.05, Table 8). Neither South Tuolumne

(before low, after low) nor Middle Tuolumne (before high, after high) exhibited a significant change (Table 8). Benthic invertebrate density, % large and small wood, and

D50 did not change at either the control or impact reaches in the first year following the

Rim Fire. Frog Creek, which represented the after-control for high-severity, exhibited a

50% increase in overhanging vegetation (p < 0.05, Table 8).

Discussion

Because of their reliance on streamside habitat and aquatic food resources, we used spiders of the family Tetragnathidae as a model organism to test the effects of fire on the ecological coupling between stream and riparian ecosystems. Although spiders operate at the local scale, we anticipated that they would integrate fire effects across spatial extents, as local stream characteristics (e.g., stream geomorphology and aquatic food resources) can reflect catchment-level processes (Poff 1997, Polis et al. 1997,

Townsend et al. 2003). In spite of pronounced differences in the conifer canopy between study reaches paired by fire severity, we found tetragnathid spider attributes to be largely

144 invariant. Results from our model-selection analysis suggest that local food resources

(i.e., benthic macroinvertebrate density as a proxy for emergent insects), catchment-scale fire frequency, and precipitation were of quantitative importance to tetragnathid spider trophic dynamics and distribution. These findings may have important implications in the context of climate change. For example, alterations in global precipitation patterns are projected to lead to increases in the duration, intensity, and frequency of extreme climate events, with more intense and less predictable flooding, droughts, and fires (Arnell 2004,

Milly et al. 2005, Westerling et al. 2006).

From a categorical (high-severity vs. low-severity) standpoint, we found that tetragnathid spider density was significantly lower in sites burned with high-severity wildfire. This result is not supported by findings by previous investigations. For instance,

Malison and Baxter (2010), in their study from central Idaho, found that riparian spider density was two-times greater at stream reaches burned by high-severity wildfire that occurred five to six years prior to sampling. Concurrently, they observed that the flux of emergent insects into the riparian zone was three-times greater in high-severity burned reaches.

In temperate ecosystems, removal of the riparian canopy by fire can trigger an increase in autochthonous productivity (Rugenski and Minshall 2014) and subsequent secondary productivity in the form of benthic macroinvertebrate biomass (Gresswell

1999, Malison and Baxter 2010, Rugenski and Minshall 2014). This aligns with our result that benthic macroinvertebrate density was 1.5 times greater in study reaches classified as high-severity burned. However, we found that benthic macroinvertebrate density was

145 negatively correlated with both tetragnathid spider density and tetragnathid spider TP.

This finding may indicate a dietary shift by spiders towards terrestrial prey in high- severity study reaches as terrestrial invertebrate abundance has been associated with increased shrub growth following wildfire (Romero et al. 2005), although we did not measure this directly. We also observed a negative relationship between fire frequency and variability in TP (i.e., Prediction 4), suggesting that catchment-level variability in fire frequency may also mediate spider-prey relationships.

Similar to (Gillespie 1987), who reported that access to open water was the most important factor determining web placement by riparian orb-web spiders, Tagwireyi and

Sullivan (2015) found that web-building substrate was more predictive than aquatic insect abundance for density of shoreline tetragnathid spiders, and this may be the case in our system where we observed no differences in overhanging vegetation between burn treatments [although we observed an increase in % overhanging vegetation and a decrease in tetragnathid spider density post Rim fire at Frog Creek (Table 6)]. Large and small wood may also provide web-building substrate for riparian orb-weaving spiders

(Gillespie 1987), and we observed a greater percentage of each study reach occupied by large and small wood in high-severity burned reaches, but we did not find support for a relationship between wood and spider density in either our paired or model-selection analyses. The occurrence of large wood in streams following wildfires may be site specific and interact strongly with time [i.e., over long timescales following stand- replacing fire, large wood delivery may be low until forest structure (e.g., trees of sufficient age and size) returns to pre-fire condition (Gresswell 1999)]. Although we did

146 observe ~30% more wood in high-severity burned sites, variability was also relatively high (SD = 47.9%), which may have contributed to our inability to detect a relationship between spiders and wood.

Fire frequency emerged as a predictor for spider trophic models, suggesting that the occurrence rate of fire (i.e., interval between fires) may affect subsidy dynamics between terrestrial and aquatic systems. Fire-frequency has been predicted to increase under various climate change scenarios (Westerling et al. 2006, Westerling et al. 2011), and therefore, the potential association between fire frequency and FCL (as measured by tetragnathid spider TP in our study) may warrant further study.

The influence of precipitation on multiple characteristics of tetragnathid spiders suggests environmental variability in climate may have a greater influence on shoreline spiders than the effects of wildfire. Precipitation emerged as a key variable for density of tetragnathid spiders (positive relationship) and was weakly associated with both spider reliance on aquatically-derived energy (negative relationship) and Hg body loading

(positive relationship), which provides initial evidence that climatic variability among our study catchments (and between study years) may exert a greater impact on stream biotic responses than fire. This conclusion was supported by post-Rim fire observations: before- after tetragnathid spider densities were different between years in the control reaches

(Frog and Cascade Creeks, Table 7) but not between years in the impact reaches (Middle and South Tuolumne, Table 8). Both direct and indirect effects are likely at play (e.g.,

Warren and Liss 1980).

147

Shifts in the distribution and composition of stream invertebrate food resources are expected to have profound effects on riparian consumers, including tetragnathid spiders (Kennedy and Turner 2011) and precipitation patterns can influence benthic invertebrate density and community composition. In a long-term climate study conducted in the Sierra Nevada, Bêche and Resh (2007) found that benthic invertebrate communities shifted from “dry-year” communities adapted to drought conditions to “wet- year” communities adapted to high flows. Small streams exhibited greater shifts due to variability in precipitation than larger streams and small stream invertebrate communities were more stable under drought conditions than larger stream communities. These shifts were largely driven by changes in the abundance of Chironomidae larvae, which may be a key food source for riparian spiders (Alp et al. 2013). Extreme precipitation events have also been shown to prompt shifts in the richness and evenness of macroinvertebrate assemblages in both the Mediterranean Basin of southern Portugal (Feio et al. 2010) and in northern California (Filipe et al. 2013) and Lawrence et al. (2010) found that longer- lived and larger benthic macroinvertebrates were less abundant in years with extreme temperature and precipitation.

In northeastern Spain, Verkaik et al. (2013b) observed that fire combined with dry-wet cycles explained significantly more variation in benthic macroinvertebrate community composition than fire alone and that richness was 30% greater in wet years compared to dry years, concluding that extreme drought resulting in cessation of flow was the primary driver of benthic macroinvertebrate community composition in their study system. In addition, in a long-term study conducted in the Big Creek basin of

148 central Idaho, Rugenski and Minshall (2014) found that climate-driven shifts towards higher temperatures and an absence of scouring flows due to reduced snowpack was associated with a reduction in the magnitude of wildfire effects on stream primary productivity and benthic macroinvertebrate community composition. Streams that had been burned by high-severity fire exhibited increased primary and secondary productivity

(chlorophyll-a and benthic invertebrate biomass, respectively) due to increased nutrient loading and removal of the conifer canopy. However, irrespective of fire, changes to stream flow, including altered timing and reduced magnitude of peak flows, have also been shown to result in higher concentrations of chlorophyll-a and standing crop of periphyton (Davis et al. 2013). Further, Arkle et al. (2010) found that burned streams exhibited greater interannual variability in benthic macroinvertebrate community composition and that year to year variability was primarily driven by streamflow.

Similarly, we found preliminary support that fire, precipitation, and benthic macroinvertebrate density interact to influence tetragnathid spider distribution and food- web characteristics in our study system; however, we did not measure stream flow directly.

This study occurred at a time of worsening drought conditions in California’s

Sierra Nevada Range. Winter snowpack in the 2013-2014 water year was 17% of average, and annual precipitation accrual for Yosemite was 45% of the long-term average

(California Data Exchange Center, 2014). Based on the significant effect of precipitation across multiple spider responses observed in the pre-Rim Fire component of this study, drought may at least in part explain the decrease in tetragnathid spider density in two of

149 the four sites surveyed in 2014 following the Rim Fire. Because wildfires in

Mediterranean ecosystems can be especially heterogeneous over space and time (Keeley et al. 2009), idiosyncratic patterns in aquatic-terrestrial responses to fire severity, frequency, and extent may be common and could contribute to the surprising lack of relationships observed from our paired analysis (both pre- and post-Rim Fire).

Although our findings indicate that local riparian consumers in our study system may be more heavily influenced by overarching patterns in precipitation than wildfire, it is likely that integrated explanations that involve fire, climate, and their interactions will be most compelling (e.g. Arkle et al. 2010, Verkaik et al. 2013b, Rugenski and Minshall

2014) especially given seasonal dry-wet cycles and high interannual variability in precipitation characteristic of Mediterranean biomes. Additionally, as fire can span multiple spatial extents within a catchment, more explicit treatment of the relative effects of local- and landscape-scale fire characteristics will be critical. This will be particularly important in investigating the effects of fire on stream-riparian food webs that include more mobile consumers (e.g., bats, birds, lizards) that might be expected to more broadly integrate stream-riparian resources and habitats. Because precipitation patterns as well as wildfire severity, frequency, and extent are projected to change dramatically under future climate change scenarios (Miller and Urban 1999, Westerling et al. 2006, Miller et al.

2009), future work examining linked fire-precipitation impacts on aquatic-terrestrial responses will be of considerable conservation and management benefit.

Acknowledgements

150

Funding was provided by NSF DEB-1401480 awarded to SMPS and BKJ, Bureau of

Land Management (14-3-01-37) award to SMPS, and The Ohio State University. In

Yosemite National Park we received assistance from Dr. G. Smith and K. Van

Wagtendonk. We appreciate field and laboratory help received from D. Groff, M.

Hickson, L. Meyer, M. Ledford, and D. Vent.

151

Literature Cited

Agee, J. K. 1993. Fire ecology of the Pacific Northwest forests. Island Press, Washington, D.C. Alberts, J. M., S. M. P. Sullivan, and A. Kautza. 2013. Riparian swallows as integrators of landscape change in a multiuse river system: implications for aquatic-to- terrestrial transfers of contaminants. Science of the Total Environment 463:42-50. Allen, D. C. and C. C. Vaughn. 2010. Complex hydraulic and substrate variables limit freshwater mussel species richness and abundance. Journal of the North American Benthological Society 29:383-394. Alp, M., B. L. Peckarsky, S. M. Bernasconi, and C. T. Robinson. 2013. Shifts in isotopic signatures of animals with complex life-cycles can complicate conclusions on cross-boundary trophic links. Aquatic Sciences 75:595-606. Anderson, D. R. and K. P. Burnham. 2002. Avoiding pitfalls when using information- theoretic methods. Journal of Wildlife Management 66:912-918. Arkle, R. S. and D. S. Pilliod. 2010. Prescribed fires as ecological surrogates for wildfires: a stream and riparian perspective. Forest Ecology and Management 259:893-903. Arkle, R. S., D. S. Pilliod, and K. Strickler. 2010. Fire, flow and dynamic equilibrium in stream macroinvertebrate communities. Freshwater Biology 55:299-314. Arnell, N. W. 2004. Climate change and global water resources: SRES emissions and socio-economic scenarios. Global Environmental Change-Human and Policy Dimensions 14:31-52. Barbour, M. T., J. Gerritsen, B. D. Snyder, and J. B. Stribling. 1999. Rapid bioassessment protocols for use in streams and wadeable rivers: periphyton, benthic macroinvertebrates and fish. U.S. Environmental Protection Agency; Office of Water, Washington, D.C. Baxter, C. V., K. D. Fausch, and W. C. Sounders. 2005. Tangled webs: reciprocal flows of invertebrate prey link streams and riparian zones. Freshwater Biology 50:20. Bêche, L. A. and V. H. Resh. 2007. Short-term climatic trends affect the temporal variability of macroinvertebrates in California 'Mediterranean' streams. Freshwater Biology 52:2317-2339. Bêche, L. A., S. L. Stephens, and V. H. Resh. 2005. Effects of prescribed fire on a Sierra Nevada (California, USA) stream and its riparian zone. Forest Ecology and Management 218:37-59. Bendix, J. and C. M. Cowell. 2010. Fire, floods and woody debris: Interactions between biotic and geomorphic processes. Geomorphology 116:297-304.

152

Blais, J. M. 2005. Biogeochemistry of persistent bioaccumulative toxicants: processes affecting the transport of contaminants to remote areas. Canadian Journal of Fisheries and Aquatic Sciences 62:236-243. Bonada, N. and Resh, V.H. 2013. Mediterranena-climate streams and rivers: geographically separated but ecologically comparable freshwater systems. Hydrobiologia 719: 1-29. Burdon, F. J. and J. S. Harding. 2008. The linkage between riparian predators and aquatic insects across a stream-resource spectrum. Freshwater Biology 53:330-346. Burnham, K. P. and D. R. Anderson. 2004. Multimodel inference - understanding AIC and BIC in model selection. Sociological Methods & Research 33:261-304. Carrion, J. S., P. Sanchez-Gomez, J. F. Mota, R. Yll, and C. Chain. 2003. Holocene vegetation dynamics, fire and grazing in the Sierra de Gador, southern Spain. Holocene 13:839-849. Cianfrani, C. M., S. M. P. Sullivan, W. C. Hession, and M. C. Watzin. 2009. Mixed stream channel morphologies: implications for fish community diversity. Aquatic Conservation: Marine and Freshwater Ecosystems 19:147-156. Covich, A. P., M. A. Palmer, and T. A. Crowl. 1999. The role of benthic invertebrate species in freshwater ecosystems. BioScience 49:119-127. Davis, F. W., E. A. Keller, A. Parikh, and J. Florsheim. 1989. Recovery of the chaparral riparian zone after wildfire. Pages 194-203 in D. L. Abell, editor. Proceedings of the California Riparian Systems Conference: Protection, Management, and Restoration for the 1990s. USDA General Technical Report Pacific Southwest, Davis, CA. Davis, J. M., C. V. Baxter, E. J. Rosi-Marshall, J. L. Pierce, and B. T. Crosby. 2013. Anticipating stream ecosystem responses to climate change: toward predictions that incorporate effects via land-water linkages. Ecosystems 16:909-922. Downes, B. J., L. A. Marmuta, P. G. Fairweather, D. P. Faith, M. J. Keough, P. S. Lake, B. D. Mapstone, and G. P. Quinn. 2002. Monitoring ecological impacts: concepts and practice in flowing waters. Cambridge University Press, Cambridge, UK. Dwire, K. A. and J. B. Kauffman. 2003. Fire and riparian ecosystems in landscapes of the western USA. Forest Ecology and Management 178:61-73. Fausch, K. D., Y. Taniguchi, S. Nakano, G. D. Grossman, and C. R. Townsend. 2001. Flood disturbance regimes influence rainbow trout invasion success among five holarctic regions. Ecological Applications 11:1438-1455. Feio, M. J., C. N. Coimbra, M. A. S. Graca, S. J. Nichols, and R. H. Norris. 2010. The influence of extreme climatic events and human disturbance on macroinvertebrate community patterns of a Mediterranean stream over 15 years. Journal of the North American Benthological Society 29:1397-1409.

153

Filipe, A. F., J. E. Lawrence, and N. Bonada. 2013. Vulnerability of stream biota to climate change in mediterranean climate regions: a synthesis of ecological responses and conservation challenges. Hydrobiologia 719:331-351. Flannigan, M., B. Stocks, M. Turetsky, and M. Wotton. 2009. Impacts of climate change on fire activity and fire management in the circumboreal forest. Global Change Biology 15:549-560. Frissell, C. A., W. J. Liss, C. E. Warren, and M. D. Hurley. 1986. A hierarchical framework for stream habitat classification: viewing streams in a watershed context. Environmental Management 10:199-215. Gillespie, R. G. 1987. The mechanism of habitat selection in the long-jawed orb-weaving spider Tetragnatha elongata (Araneae, Tetragnathidae). Journal of Arachnology 15:81-90. Gresswell, R. E. 1999. Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193-221. Grove, A.T. and Rackham, O. 2001. The nature of Mediterranean Europe: an ecological history. Journal of Mediterranean Ecology 3: 65-68. Henschel, J. R., D. Mahsberg, and H. Stumpf. 2001. Allochthonous aquatic insects increase predation and decrease herbivory in river shore food webs. Oikos 93:429-438. Hynes, H. B. N. 1975. The stream and its valley. Verh. Internat. Verein. Limnol. 19:1-15. Jackson, B. K. and S. M. P. Sullivan. 2009. Influence of fire severity on riparian vegetation heterogeneity in an Idaho, U.S.A. wilderness. Forest Ecology and Management 259:24-32. Jackson, B. K., S. M. P. Sullivan, and R. L. Malison. 2012. Wildfire severity mediates fluxes of plant material and terrestrial invertebrates to mountain streams. Forest Ecology and Management 278:27-34. Keeley, J. E., H. Safford, C. J. Fotheringham, J. Franklin, and M. Moritz. 2009. The 2007 southern California wildfires: lessons in complexity. Journal of Forestry 107:287- 296. Kennedy, T. L. and T. F. Turner. 2011. River channelization reduces nutrient flow and macroinvertebrate diversity at the aquatic terrestrial transition zone. Ecosphere 2:1-13. Key, C.H. and N.C. Benson. 2006. Landscape assessment: ground measure of severity, the Composite Burn Index; and remote sensing of severity, the Normalized Burn Ratio. In D.C. Lutes; R.E. Keane; J.F. Caratti; C.H. Key; N.C. Benson; S. Sutherland; and L.J. Gangi. 2006. FIREMON: Fire effects monitoring and inventory system. USDA Forest Service, Rocky Mountain Research Station, Ogden, UT. Gen. Tech. Rep. RMRS-GTR-164-CD: LA 1-51.

154

Koetsier, P., T. R. B. Krause, and Q. M. Tuckett. 2010. Present effects of past wildfires on leaf litter breakdown in stream ecosystems. Western North American Naturalist 70:164-174. Lawrence, J. E., K. B. Lunde, R. D. Mazor, L. A. Beche, E. P. McElravy, and V. H. Resh. 2010. Long-term macroinvertebrate responses to climate change: implications for biological assessment in mediterranean-climate streams. Journal of the North American Benthological Society 29:1424-1440. Legleiter, C. J., R. L. Lawrence, M. A. Fonstad, W. A. Marcus, and R. Aspinall. 2003. Fluvial response a decade after wildfire in the northern Yellowstone ecosystem: a spatially explicit analysis. Geomorphology 54:119-136. Lenihan, J. M., R. Drapek, D. Bachelet, and R. P. Neilson. 2003. Climate change effects on vegetation distribution, carbon, and fire in California. Ecological Applications 13:1667-1681. Malison, R. L. and C. V. Baxter. 2010. The fire pulse: wildfire stimulates flux of aquatic prey to terrestrial habitats driving increases in riparian consumers. Canadian Journal of Fisheries and Aquatic Sciences 67:570-579. May, C. L. and R. E. Gresswell. 2003. Processes and rates of sediment and wood accumulation in headwater streams of the Oregon Coast Range, USA. Earth Surface Processes and Landforms 28:409-424. McHugh, P. A., A. R. McIntosh, and P. G. Jellyman. 2010. Dual influences of ecosystem size and disturbance on food chain length in streams. Ecology Letters 13:881-890. Mellon, C. D., M. S. Wipfli, and J. L. Li. 2008. Effects of forest fire on headwater stream macroinvertebrate communities in eastern Washington, USA. Freshwater Biology 53:2331-2343. Meyer, L. A. and S. M. P. Sullivan. 2013. Bright lights, big city: influences of ecological light pollution on reciprocal stream-riparian invertebrate fluxes. Ecological Applications 23:1322-1330. Mihuc, T. B. and G. W. Minshall. 1995. Trophic generalists vs trophic specialists - implications for food-web dynamics in post-fire streams. Ecology 76:2361-2372. Mihuc, T. B. and G. W. Minshall. 2005. The trophic basis of reference and post-fire stream food webs 10 years after wildfire in Yellowstone National Park. Aquatic Sciences 67:541-548. Miller, C. and D. L. Urban. 1999. Forest pattern, fire, and climatic change in the Sierra Nevada. Ecosystems 2:76-87. Miller, J. D., H. D. Safford, M. Crimmins, and A. E. Thode. 2009. Quantitative evidence for increasing forest fire severity in the Sierra Nevada and Southern Cascade mountains, California and Nevada, USA. Ecosystems 12:16-32. Milly, P. C. D., K. A. Dunne, and A. V. Vecchia. 2005. Global pattern of trends in streamflow and water availability in a changing climate. Nature 438:347-350. 155

Montgomery, D. R. and J. M. Buffington. 1997. Channel-reach morphology in mountain drainage basins. GSA Bulletin 109:596-611. Moritz, M. A., M.-A. Parisien, E. Batllori, M. A. Krawchuk, J. Van Dorn, D. J. Ganz, and K. Hayhoe. 2012. Climate change and disruptions to global fire activity. Ecosphere 3. Murakami, M. and S. Nakano. 2002. Indirect effect of aquatic insect emergence on a terrestrial insect population through bird predation. Ecology Letters 5:333. Parker, M. S. 1994. Feeding ecology of stream-dwelling Pacific giant salamander larvae (Dicamptodon tenebrosus). Copeia 1994:705-718. Parnell, A., R. Inger, S. Bearhop, and A. Jackson. 2010. Coping with too much variation. PLoS ONE 5:e9672. Parnell, A. and A. Jackson. 2013. Stable Isotope Analysis in R. http://cran.r- project.org/web/pakage/siar/. Pimm, S. L. 2002. Food webs. The University of Chicago Press, Chicago. Poff, L. N. 1997. Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16:391-409. Poff, N. L. and J. V. Ward. 1990. Physical habitat template of lotic systems - recovery in the context of historical pattern of spatiotemporal heterogeneity. Environmental Management 14:629-645. Polis, G. A., W. B. Anderson, and R. D. Holt. 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Review of Ecology and Systematics 28:289-316. Post, D. M. 2002. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83:703-718. Power, M. E. and W. E. Rainey. 2000. Food webs and resource sheds: towards spatially delimiting trophic interactions. Pages 291-314 in M. J. Hutchings, E. A. John, and A. J. Stewart, editors. The ecological consequences of environmental heterogeneity. Blackwell Science, London, England. Power, M. E., W. E. Rainey, M. S. Parker, J. L. Sabo, A. Smyth, S. Khandwala, J. C. Finlay, F. C. McNeely, K. Marsee, and C. Anderson. 2004. River to watershed subsidies in an old-growth conifer forest. Pages 217-240 in G. A. Polis, M. E. Power, and G. R. Huxel, editors. Food webs at the landscape level. The University of Chicago Press, Chicago, IL, USA. Romero, N., R. E. Gresswell, and J. L. Li. 2005. Changing patterns in coastal cutthroat trout (Oncorhynchus clarki clarki) diet and prey in a gradient of deciduous canopies. Canadian Journal of Fisheries and Aquatic Sciences 62:1797-1807.

156

Rowse, L. M., A. D. Rodewald, and S. M. P. Sullivan. 2014. Pathways and consequences of contaminant flux to Acadian flycatchers (Empidonax virescens) in urbanizing landscapes of Ohio, USA. Science of the Total Environment 485:461-467. Rugenski, A. T. and G. W. Minshall. 2014. Climate-moderated responses to wildfire by macroinvertebrates and basal food resources in montane wilderness streams. Ecosphere 5. Sabo, J. L., J. C. Finlay, T. Kennedy, and D. M. Post. 2010. The role of discharge variation in scaling of drainage area and food chain length in rivers. Science 330:965-967. Sabo, J. L., J. C. Finlay, and D. M. Post. 2009. Food chains in freshwaters. Pages 187- 220 in R. S. Ostfeld and W. H. Schlesinger, editors. Year in Ecology and Conservation Biology 2009. Sanzone, D. M., J. L. Meyer, E. Marti, E. P. Gardiner, J. L. Tank, and N. B. Grimm. 2003. Carbon and nitrogen transfer from a desert stream to riparian predators. Oecologia 134:238-250. Shakesby, R. A. 2011. Post-wildfire soil erosion in the Mediterranean: review and future research directions. Earth-Science Reviews 105:71-100. Spencer, C. N., K. O. Gabel, and F. R. Hauer. 2003. Wildfire effects on stream food webs and nutrient dynamics in Glacier National Park, USA. Forest Ecology and Management 178:141-153. Stephens, S. L., R. E. Martin, and N. E. Clinton. 2007. Prehistoric fire area and emissions from California's forests, woodlands, shrublands, and grasslands. Forest Ecology and Management 251:205-216. Stewart-Oaten, A., W. W. Murdoch, and K. R. Parker. 1986. Environmental-impact assessment - psuedoreplication in time. Ecology 67:929-940. Sullivan, S. M. P. and A. D. Rodewald. 2012. In a state of flux: the energetic pathways that move contaminant from aquatic to terrestrial environments. Environmental Toxicology and Chemistry 31:1175-1183. Sullivan, S. M. P., M. C. Watzin, and W. S. Keeton. 2007. A riverscape perspective on habitat associations among riverine bird assemblages in the Lake Champlain Basin, USA. Landscape Ecology 22:1169-1186. Tagwireyi, T. and S. M. P. Sullivan. 2015. Distribution and trophic dynamics of riparian tetregnathid spiders in a large river system. Marine and Freshwater Research. Townsend, C. R., S. Doledec, R. Norris, K. Peacock, and C. Arbuckle. 2003. The influence of scale and geography on relationships between stream community composition and landscape variables: description and prediction. Freshwater Biology 48:768-785.

157

Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Canadian Journal of Fishery and Aquatic Science 37:130-137. Vaz, P. G., D. R. Warren, P. Pinto, E. C. Merten, C. T. Robinson, and F. C. Rego. 2011. Tree type and forest management effects on the structure of stream wood following wildfires. Forest Ecology and Management 262:561-570. Verkaik, I., M. Rieradevall, S. D. Cooper, J. M. Melack, T. L. Dudley, and N. Prat. 2013a. Fire as a disturbance in mediterranean climate streams. Hydrobiologia 719:353-382. Verkaik, I., M. Vila-Escale, M. Rieradevall, and N. Prat. 2013b. Seasonal drought plays a stronger role than wildfire in shaping macroinvertebrate communities of Mediterranean streams. International Review of Hydrobiology 98:271-283. Vieira, N. K. M., T. R. Barnes, and K. A. Mitchell. 2011. Effects of wildfire and postfire floods on stonefly detrivores of the Pajarito Plateau, New Mexico. Western North American Naturalist 71:257-270. Voshell, J. R. 2002. A guide to common freshwater invertebrates of North America. The McDonald and Woodward Publishing Company, Blacksburg, Virginia. Walters, D. M., K. M. Fritz, and R. R. Otter. 2008. The dark side of subsidies; adult stream insects export organic contaminants to riparian predators. Ecological Applications 18:1835-1841. Walters, D. M., M. A. Mills, K. M. Fritz, and D. F. Raikow. 2010. Spider-mediated flux of PCBs from contaminated sediments to terrestrial ecosystems and potential risks to arachnivorous birds. Environmental Science & Technology 44:2849-2856. Warren, C. E. and W. J. Liss. 1980. Adaptation to aquatic environments. Pages 15-40 in R. T. Lackey and L. Nielsen, editors. Fisheries management. Blackwell Scientific Publications, Oxford. Westerling, A. L., B. P. Bryant, H. K. Preisler, T. P. Holmes, H. G. Hidalgo, T. Das, and S. R. Shrestha. 2011. Climate change and growth scenarios for California wildfire. Climatic Change 109:445-463. Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and earlier spring increase western US forest wildfire activity. Science 313:940-943. Williams, D. D., L. G. Ambrose, and L. N. Browning. 1995. Trophic dynamic of two sympatric species of riparian spider (Araneae, Tetragnathidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie 73:1545-1553. Wolman, M. G. 1954. A method of sampling coarse river-bed material. Transactions of the American Geophysical Union 35:951-956.

158

Table 5. Stream reaches from Yosemite National Park were paired by fire severity with an attempt to minimize variation in time

since fire, elevation, aspect, stream channel morphology, and dominant vegetation. “Year burned” refers to the last time the study

reach burned; in some cases there was a more recent fire elsewhere in the catchment. Channel-reach morphology was classified

following Montgomery and Buffington (1997). Stream order was determined based on Strahler (1957).

Paired reaches Burn Year Elevation Aspect Order Study-reach Channel-reach severity burned (m) length (m) morphology Buena Vista High 2001 2154 NNW 3 80.3 Step-pool

5 Mono Low 2004 2096 N 3 62.5 Pool-riffle

Frog High 2006 1909 S 3 95.8 Step- pool Cascade Low 2007 1864 SW 3 75.7 Step-pool

Middle Tuolumne High 1996 1766 SW 3 125.0 Pool- riffle South Tuolumne Low 2002 1716 NW 3 139.5 Step-pool

Continued b

159

Table 5 continued

Tamarack High 2009 1931 S 2 43.2 Step-pool Grouse Low 2009 1574 NW 3 60.7 Step-pool

Crane High 2009 1850 E 2 19.7 Step- pool Coyote Low 2007 1836 SE 2 47.8 Step-pool

Meadow High 2005 2127 N 1 13.1 Pool- riffle

6 Camp Low 2005 2098 N 1 21.5 Pool-riffle

b Continued

160

Table 5 continued

Paired reaches Dominant riparian vegetation Dominant upland vegetation

Buena Vista Salix, Rubus, Alnus Subalpine (Lodgepole Pine) Mono Salix, Cornus, Alnus Subalpine (Lodgepole Pine)

Frog Salix, Cornus, Rhododendron Lodgepole Pine and Red Fir Cascade Salix, Cornus, Rhododendron Lodgepole Pine and Red Fir

Middle Tuolumne Salix, Populus, Rhododendron Yellow Pine Belt

6 South Tuolumne Salix, Alnus Yellow Pine Belt

Tamarack Salix Lodgepole Pine and Red Fir Grouse Alnus, Cornus Yellow Pine Belt

Crane Cornus, Rhododendron Lodgepole Pine and Red Fir Coyote Salix, Cornus, Rhododendron Lodgepole Pine and Red Fir

Continued b

161

Table 5 continued

Meadow Salix Subalpine (Lodgepole Pine) Camp Alnus, Cornus Subalpine (Lodgepole Pine)

162

Table 6. Data for each study reach (by low-severity and high-severity fire pair) for tetragnathid spider responses, benthic

macroinvertebrates, shoreline habitat, stream geomorphology, precipitation, catchment size, and fire frequency (proportion of

catchment burned > 2x since 1930) and fire extent (proportion of catchment burned with moderate-to-high severity fire) at the 12

paired reaches surveyed in 2011 and 2012. TP is trophic position. SD is standard deviation. Hg is mercury concentration. EPT

indicates percent of the benthic macroinvertebrate community from the orders Ephemeroptera, Plecoptera, and Trichoptera. D50 is

median sediment size. Precipitation (sampling year indicated in parentheses) was calculated as the average of each monthly total

between 01 Oct of the previous year to 30 Sept of the year sampling occurred.

Tetragnathid spider responses Paired reaches Burn Tetragnathid Reliance on TP SD of TP [Hg] severity density aquatically-derived energy

no. m-1 % ppb Buena Vista High 1.4 35.2 2.58 0.59 386 Mono Low 3.5 19.9 3.04 0.70 604

Frog High 3.0 46.9 2.64 0.63 406 Cascade Low 1.2 51.7 3.22 0.83 324

b Continued 163

Table 6 continued

Middle Tuolumne High 1.8 49.0 2.86 0.69 447 South Tuolumne Low 3.2 45.6 2.69 0.70 505

Tamarack High 3.8 47.4 2.71 0.62 347 Grouse Low 6.2 46.0 2.59 0.60 427

Crane High 1.9 56.5 2.08 0.59 213

1 Coyote Low 1.7 46.0 2.44 0.59 328

4 Meadow High 1.0 * * * * Camp Low 3.5 48.2 2.46 0.57 *

Continued

164

Table 6 continued

Macroinvertebrates Riparian structure Paired reaches Benthic EPT Burn Overhanging Wood - invertebrate severity vegetation large and density small

no. m-2 % % % Buena Vista 58.4 85.4 High 91.9 33.9 Mono 29.6 65.4 Low 53.2 12.9

Frog 42.4 83.5 High 100.0 101.6

5 Cascade 21.7 59.9 Low 83.9 21.0

Middle Tuolumne 23.6 66.0 High 100.0 66.7 South Tuolumne 36.0 77.8 Low 77.4 51.6

Tamarack 30.0 58.1 High 95.2 138.7 Grouse 24.7 82.4 Low 85.5 62.9

Crane 54.2 77.8 High 40.3 3.2 Coyote 35.3 61.3 Low 100.0 21.0

Continued b

165

Table 6 continued

Meadow 66.8 60.5 High 100.0 67.7 Camp 30.0 84.4 Low 58.1 54.8

Continued

166

Table 6 continued

Geomorphology Catchment-scale variables Paired reaches Width Depth Width-to- D50 Precipitation Catchment Fire Fire depth ratio size Frequency Extent

m m mm cm month-1 km2 % % Buena Vista 8.0 0.4 24.6 22.8 8.7 3524.1 14.8 24.6 Mono 6.3 0.5 13.3 11.4 8.7 2076.5 29.3 21.1

6 Frog 9.6 0.4 21.7 8.2 3.9 4388.8 41.8 36.9

7 Cascade 7.6 0.9 8.9 11.3 3.9 2688.2 1.5 1.5

Middle Tuolumne 12.5 0.7 19.1 11.3 3.9 9429.0 11.4 42.7 South Tuolumne 14.0 0.9 19.6 8.2 8.6 3477.0 1.5 9.6

Tamarack 4.3 0.9 4.7 2.9 8.6 1064.9 52.6 85.4 Grouse 6.1 0.5 11.5 11.3 13.5 1012.6 70.8 43.4

Crane 2.0 0.6 4.0 8.1 3.9 653.4 72.8 41.0 Coyote 4.8 0.6 9.4 16.1 3.9 493.5 9.5 2.4

Continued b

167

Table 6 continued

Meadow 1.3 0.3 4.1 2.1 8.7 196.3 91.3 29.0 Camp 2.2 0.5 4.7 1.8 8.7 141.7 99.6 35.8

* We were unable to collect spiders of sufficient size and number to determine trophic position (TP), reliance on aquatically-

6 derived energy, and mercury Hg at Meadow Creek and Hg at Camp Creek.

168

Table 7. Retained regression models (ΔAICc ≤ 4) with corresponding AICc scores, Akaike weights (wi), and variation explained

(R2). Null models (i.e., intercept only) are also included. See text for description of independent variables.

2 Tetragnathid response AICc ΔAICc wi R Density Precipitation (+), Benthic macroinvertebrate density (-) 41.92 0.00 0.58 0.78 Precipitation (+) 43.73 1.81 0.23 0.53 Precipitation (+), Benthic macroinvertebrate density (-), EPT (% of community 44.48 2.56 0.16 0.86 composition) (+) Null 47.75 5.83 0.03 0.00

1 Trophic position (mean)

9 Benthic macroinvertebrate density (-), Fire frequency (-) 6.30 0.00 0.50 0.66

Benthic macroinvertebrate density (-) 7.05 0.75 0.35 0.41 Null 9.00 2.70 0.13 0.00

Trophic position (SD) Fire frequency (-) -24.38 0.00 0.40 0.37 Fire frequency (-) , Benthic macroinvertebrate density (-) -24.18 0.20 0.36 0.60 Null -23.23 1.14 0.22 0.00

b Continued 169

Table 7 continued

Reliance on aquatically-derived energy Null -14.79 0.00 0.78 0.00 Precipitation (-) -11.97 2.82 0.19 0.10

Mercury (Hg) Null 126.64 0.00 0.68 0.00 Precipitation (+) 128.68 2.04 0.25 0.20

b 170

Δ t p Middle Tuolumne (before high, after high) Tetragnathidae density (no. m-1) 0.7 -1.16 0.282 Benthic macroinvertebrate density (no. m-2) 34.2 -0.81 0.434

D50 (mm) 4.8 0.32 0.767 Wood - large and small (%) -0.3 1.34 0.213 Overhanging vegetation (%) 0.2 0.53 0.610

Frog Creek (before high, after control) Tetragnathidae density (no. m-1) -2.1 3.91 0.005 Benthic macroinvertebrate density (no. m-2) -12.7 0.37 0.717

D50 (mm) 0.0 1.44 0.286 Wood - large and small (%) 0.1 -0.45 0.667 Overhanging vegetation (%) 0.5 -4.81 0.004

South Tuolumne (before low, after low) Tetragnathidae density (no. m-1) -2.1 1.42 0.118 Benthic macroinvertebrate density (no. m-2) 36.4 -1.21 0.255

D50 (mm) -0.1 0.88 0.472 Wood - large and small (%) -0.3 1.14 0.282 Overhanging vegetation (%) -0.2 1.34 0.226 Continued 171

Table 8 continued

Cascade Creek (before low, after control) Tetragnathidae density (no. m-1) -2.5 5.42 0.001 Benthic macroinvertebrate density (no. m-2) 26.8 -1.40 0.188

D50 (mm) 4.7 1.20 0.296 Wood - large and small (%) * * * Overhanging vegetation (%) -0.3 1.72 0.125

* Data unavailable due to an error in data processing.

172

Figure 14. Fire history by decade within Yosemite National Park. Paired reaches are

Tamarack (1a) and Grouse (1b); Meadow (2a) and Camp (2b); Buena Vista (3a) and

Mono (3b); Middle Tuolumne (4a) and South Tuolumne (4b); Frog (5a) and Cascade

(5b); and Crane (6a) and Coyote (6b). Weather stations are White Wolf, Crane Flat,

Wawona, and Mariposa Grove. 173

Figure 15. The 2013 Rim Fire perimeter and paired stream study reaches sampled in 2011 and 2012. Middle Tuolumne (4b) and South Tuolumne (4a) were consumed by the Rim

Fire with high- and low-severity, respectively and were used as impact study reaches in

2014. Frog (5a) and Cascade (5b) were unburned by the Rim Fire and used as high- and low-severity control reaches, respectively.

174

Figure 16. Average monthly precipitation (cm) for the Crane Flat weather station for each quarter (i.e., October to December, January to March, April to June, and July to

September) of each water year (October to September) in the study period (2010-2014).

175

(a)

density

)

- (# m (#

Tetragnathidae p = 0.195

(b)

position Trophic

- p = 0.220

(c)

aquatically derived energy derived

p = 0.500 Reliance on Reliance

(d)

/kg) /kg)

μ Hg ( Hg

p = 0.180

Asteriks represent mean and bars represent +/- one standard deviation from the mean. N

= 12 for all tests except for b (n = 11), c (n = 11), and d (n = 10), where insufficient sample was available for stable isotope and Hg analyses.

Continued 176

Figure 17 continued

(e)

large large and

– small small (%)

Wood Wood p = 0.103

(f) Overhanging vegetation(%) p = 0.478

(g)

(mm)

50 D

p = 0.621

(h) Width/depth ratio Width/depth p = 0.812

Continued

177

Figure 17 continued

(i)

)

macroinvertebrate density (#densitym

Benthic p = 0.072

(j)

EPT % %

p = 0.812

178

Hg (μg/kg)Hg

Mean Trophic Position

0.30) but became significant with the removal of Cascade Creek (black dot, solid line: p

= 0.0014, R2 = 0.79).

179

Chapter 4: Variability in invertebrate predator trophic position and reliance on aquatically-derived energy linked to ecosystem size, flood magnitude, and wildfire in a California Mediterranean-climate river system.

Breeanne K. Jackson; S. Mažeika P. Sullivan

Abstract: Although wildfire can strongly influence fluvial ecosystem structure and function, its influence on stream-riparian food-web dynamics has received minimal attention to date, especially in Mediterranean-type climates. In particular, wildfire may function as a disturbance agent that drives consumer trophic dynamics in stream ecosystems. We measured the relative effects of wildfire characteristics (frequency, timing, and magnitude), hydrologic disturbance, ecosystem size, and in-stream productivity on trophic position and reliance on aquatically-derived energy (i.e., derived from benthic algae) of/by aquatic benthic macroinvertebrate predators and riparian spiders of the family Tetragnathidae along a gradient of drainage area in two rivers on the west slope of the Sierra Nevada in California, USA. Ecosystem size (i.e., drainage area and channel width) received strong support as an environmental determinant of both trophic measures for aquatic macroinvertebrates and tetragnathid spiders, with variability in flood magnitude emerging as an important mechanism linking ecosystem size and trophic responses. Greater proportion of the landscape affected by wildfire was related to lower trophic position of predatory benthic macroinvertebrates and tetragnathid spiders.

180

Piecewise linear regression revealed significant breakpoints in tetragnathid spider trophic position and reliance on aquatically-derived energy that were related to thresholds in the proportion of the catchment influenced by fire. These non-linear relationships with wildfire may lend additional insight into the potential interactions between ecosystem size, productivity, and disturbance for determining functional attributes of stream-riparian food webs.

181

Introduction

Food-web connectivity between streams and their adjacent riparian zones has been dominated by attention to lateral food subsidies originating outside the stream including terrestrial leaf litter, woody debris, and other plant material (Vannote et al.

1980, Junk et al. 1989). Reliance of fish on terrestrial invertebrates as prey has also been demonstrated in detail (Piccolo and Wipfli 2002, Allan et al. 2003, Carpenter et al. 2005).

More recently, flows of prey into the riparian zone (i.e., emerging adult aquatic insects) and terrestrial-aquatic feedback loops have received increasing attention (Nakano et al.

1999, Power 2002, Baxter et al. 2004, Baxter et al. 2005, Paetzold et al. 2005). Multiple disturbances, including land use (Alberts et al. 2013), the introduction of invasive species

(Baxter et al. 2004), and even artificial light pollution (Meyer and Sullivan 2013), can influence aquatic-terrestrial food-web dynamics. However, wildfire – a significant disturbance agent in stream-riparian ecosystems (Gresswell 1999, Bisson et al. 2003) – has only begun to receive attention as an environmental determinant of aquatic-terrestrial trophic linkages (Spencer et al. 2003, Malison and Baxter 2010b, Jackson et al. 2012,

Jackson and Sullivan In press), yet might be expected to be a strong driver of variability in food-chain length and other trophic measures.

Food-chain length (FCL) is an important characteristic of ecological communities and a primary indicator of ecosystem function and stability (Post 2002a, Sabo et al.

2009). Realized FCL, the number of transfers of energy leading to a single species in a food web (usually a top predator), is thought to be influenced by a number of factors including productivity, ecosystem size, predator-prey interactions, and disturbance (Sabo

182 et al. 2009). The dynamic constraint hypothesis (Pimm and Lawton 1977) predicts that ecosystems affected by frequent or intense disturbance should have shorter FCL. For example, McHugh et al. (2010) found that FCL in New Zealand streams was negatively related to variability in temperature, hydrology, and geomorphology. Further, recent evidence implicates hydrologic variability in mediating the effect of ecosystem size on

FCL in rivers and streams (Sabo et al. 2010).

Multiple elements of wildfire may affect food-web structure. Removal of upland and riparian vegetation and other obstructions to overland flow by wildfire can lead to increased runoff, soil erosion, deposition of sediments in streams (Wondzell and King

2003, Shakesby and Doerr 2006), and debris flows (May 2007). Wildfire-induced changes in stream hydrogeomorphology can last for years (Miller et al. 2003) and lead to simplification of community composition of benthic invertebrates (Minshall 2003, Vieira et al. 2004), which may lower trophic position of consumers by removing omnivores or intermediary predators (Marks et al. 2000). Conversely, increased stream temperatures resulting from a combination of removal of the riparian canopy (Pettit and Naiman 2007) and reorganization of the streambed (Royer and Minshall 1997, Dunham et al. 2007), can lead to greater in-stream primary and secondary productivity (Robinson et al. 1994,

Malison and Baxter 2010b, Cooper et al. 2014) which may, in turn, lead to longer FCL

(Yodzis 1984). For example, Spencer et al. (2003) found that benthic invertebrates and fish from streams that had burned five years previous were significantly more enriched in

15N, hinting at the possibility that fire may lengthen FCL in contrast to hydrologic

183 disturbance (but note that enrichment is not conclusive evidence of increased FCL by itself as other factors can lead to increased δ15N; reviewed in Wan et al. 2001).

The River Continuum Concept (Vannote et al. 1980) predicts that allochthony

(i.e., inputs to stream food webs from terrestrial organic material) will dominate in headwater streams and gradually be replaced by food webs more reliant on autochthonous primary production (i.e., from in stream autotrophy) in mid-order streams.

However, in small streams (i.e., 1st to 2nd order), wildfire has been shown to prompt an increase in authochthonous-based aquatic food webs (Rugenski and Minshall 2014) and shift benthic invertebrate community composition toward r-selected generalist taxa

(Minshall et al. 2001, Malison and Baxter 2010a), increasing nutritional prey subsidies from streams to riparian zones in the form of emergent aquatic insects (Malison and

Baxter 2010b, Jackson et al. 2012). Whether these wildfire-induced changes in the relative autochthony of stream food webs interrupt down-watershed gradients or act synergistically to create novel food web architecture has yet to be examined.

Wildfire can also interact with flow characteristics to influence stream-riparian ecosystems (Gresswell 1999, Arkle et al. 2010). In the absence of high-magnitude floods, shifts in benthic invertebrate abundance and community composition that typically follow wildfire may not occur (Arkle et al. 2010, Rugenski and Minshall 2014). Further, in drought-prone ecosystems, stream drying may have a greater effect on benthic invertebrate assemblages than wildfire (Verkaik et al. 2013). Food-web characteristics may be particularly dynamic in regions influenced by Mediterranean-type climate characterized by annual dry-wet cycles and high-interannual variability in precipitation

184 and hydrology (Bonada and Resh 2013). However, the interaction between floods, droughts, and wildfire has not been investigated within the context of stream-riparian food webs (but see Arkle et al. 2010, Verkaik et al. 2013).

Naturally-abundant stable-isotope analysis is a valuable tool for describing food webs in aquatic ecosystems because it addresses trophic position and diet (Collier et al.

2002, Hicks et al. 2005). The ratio of 13C to 12C (δ13C) can vary between terrestrial and aquatic primary producers. For example, stream algae can exhibit a distinct δ13C from riparian deciduous shrubs (Finlay 2001). This distinction is retained in consumer organisms so that the source of a consumer’s energy requirements (or diet) can be determined from the isotopic signature. In addition, the trophic position of a consumer organism can be determined from its nitrogen isotope signature as there is in general a 3-

4‰ enrichment of δ15N (the ratio of 15N to 14N) with each trophic step. Although stable isotope analysis has been widely used to examine aquatic food webs (Finlay 2001, Post

2002b) this is one of the first attempts to describe wildfire effects on aquatic-terrestrial food-web linkages using this method.

Within this context, we sought to determine the comparative influence of wildfire on the reliance on aquatically-derived energy (i.e., derived from benthic algal pathways) and trophic position of predatory benthic macroinvertebrates and riparian spiders of the family Tetragnathidae along the Merced River and South Fork of the Merced River within Yosemite National Park in central California, USA. We used trophic position of top invertebrate consumers (i.e., predatory benthic macroinvertebrates and tetragnathid spiders) as a measure of food-web architecture, which should respond to the same

185 processes that account for variation in FCL (i.e., ecosystem size, disturbance, and productivity; given that FCL is a measure of the maximum trophic position among ensembles of predators) (Post 2002a).

We hypothesized that stream reaches affected by wildfire would be discontinuous within upstream-downstream gradients in drainage area size with respect to reliance on aquatically-derived energy and trophic position of both in-stream and riparian invertebrate consumers. Specifically, we predicted that (1) reliance on aquatically- derived energy by predatory benthic macroinvertebrates and tetragnathid spiders would increase as the relative percentage of the catchment affected by frequent, recent, and/or severe wildfire increased due to opening of the channel canopy and increased light penetration (2) trophic position of predatory benthic macroinvertebrates and tetragnathid spiders would be lower relative to the proportion of the catchment burned by wildfire owing to reduced complexity in benthic invertebrate community composition; and (3) irrespective of the proportion of each catchment affected by wildfire, ecosystem size (as measured by drainage area and channel width) would be a dominant factor explaining invertebrate trophic position and the reliance on aquatically-derived energy by stream- riparian invertebrate consumers.

Methods

Yosemite National Park (YNP) is located within the central Sierra Nevada of

California, USA. The Merced River and the South Fork of the Merced River both originate on different aspects of Triple Divide Peak in the Clark Range at 2,413 m above

186 sea level. Both rivers then flow westward through glacially-carved valleys eventually joining outside the park boundary. The Merced River drains 4,470 km2 and has an average discharge of 34 m3 s-1 at its mouth and the South Fork of the Merced River drains

280 km2 and has an average discharge of 10 m3 s-1. The regional climate is

Mediterranean-type characterized by dry-wet seasonality and high interannual variability in precipitation that is heavily influenced by the El Niño Southern Oscillation cycle

(DeFlorio et al. 2013). Average annual precipitation in YNP is 94.5 cm, 73.7 cm of which falls as snow (Western Regional Climate Center, 2012).

We sampled at 31 stream reaches (~100 m) within eight valley segments –

Merced River Above Nevada Fall, Yosemite Valley, Merced Gorge, El Portal, South

Fork Merced River Above Wawona, Wawona Impoundment, Wawona, South Fork

Merced River Below Wawona –as designated in the Merced Wild and Scenic River

Values Draft Baseline Conditions Report (2011) that were distributed along gradients of elevation and drainage area (Figure 19).

At each study reach, we collected riparian spiders of the family Tetragnathidae (3-

6 individuals per reach) and predatory benthic macroinvertebrates (e.g., Megaloptera,

Plecoptera; 2-10 individuals per reach depending on size of each individual) for stable- isotope analysis of 13C and 15N. Each search effort lasted for multiple hours and we made

2-3 search efforts per study reach. However, in some cases, no benthic predatory macroinvertebrates were collected. Benthic invertebrates were identified to order in the field using Voshell (2002) as a guide. We also collected epilithic/benthic algae and stream-conditioned leaf litter (i.e., detritus) as basal resources for stable-isotope analysis.

187

We collected materials from upstream, mid-reach, and downstream locations along each sampling reach resulting in one composite sample of each basal resource per reach.

Sample processing

Tetragnathid spiders and predatory benthic macroinvertebrates were rinsed with distilled water and oven-dried at 60°C for 48 hours in the laboratory. We homogenized the tissue from 1-5 benthic maroinvertebrates and then 3-5 individual spiders using a mortar and pestle, then packed a small amount of each composite sample into tin capsules. We sorted epilithic algae and detritus from other materials (e.g., sediment, invertebrates, etc.) and rinsed the samples with distilled water. After oven-drying, subsamples of epilithic algae and detritus were combined per reach, respectively, and then homogenized into a fine powder using a Pica Blender Mill (Cianflone Scientific

Instruments Corporation, Pittsburgh, Pennsylvania) or mortar and pestle before packing composite samples (1 per study reach) for analysis.

Continuous flow isotope-ratio mass spectrometry (EA-IRMS) was used to determine 13C and 15N for all samples at Washington State University Stable Isotope Core

(Pullman, Washington). The results are reported in δ (‰) notation defined as: δ13C or

15 13 12 15 14 δ N = [(Rsample/Rstandard)-1] * 100 where R is C/ C or N/ N, respectively. Typical analytical precision was 0.08‰ for δ15N and 0.19‰ for δ 13C determination.

188

Estimating trophic position and reliance on aquatically-derived energy

After processing, we used the two-source food web model from Post (2002b) to estimate tetragnathid spider and predatory benthic macroinvertebrate trophic position; TP

= λ + { δc – [ δb1 * α + δb2 * (1-α)]}/ Δn where λ is the trophic position of the basal food

15 sources (i.e., 1 for primary producers); δc is the δ N signature of the consumer; δb1 and δb2 are the signatures of the two basal food sources; α is the proportion

15 of N from basal food source 1; and Δn is the enrichment in δ N per trophic level (i.e.,

3.4‰; Post 2002b). A two-end member Bayesian isotopic mixing model was used to determine the proportion of N derived from basal source 1 (i.e., α) with the R software package SIAR (Stable Isotope Analysis in R; Parnell and Jackson 2013). SIAR is equipped to handle variability in sources, consumers, and trophic fractionation factors (Parnell et al. 2010).

Epilithic algae and detritus were the two basal food source end members for all study reaches. To estimate the contribution from each basal food source to the consumer,

δ13C and δ15N data were used. Trophic fractionation factors for tetragnathid spiders and predatory benthic insects were estimated using the per trophic step fractionation in Post

(2002b) (i.e., 3.4‰ ± 0.98‰ for δ15N and 0.39‰ ± 1.3‰ for δ13C) multiplied by the estimated number of trophic transfers between the consumer and basal resources

(estimated a priori as the difference between the consumer δ15N and mean basal resource

δ15N divided by 3.4‰).

189

Fire variables

The catchment draining to each sampling reach was delineated using the watershed tool in ArcGIS 10.1 (Esri, Redlands, California) from a 10 m2 digital elevation model (DEM) of the Yosemite region acquired from the YNP fire atlas and created by the

US Geological Survey (USGS). The proportion of each catchment that had been burned in the last ten years (i.e., fire timing), greater than twice since 1930 (i.e., fire frequency), and the proportion burned with moderate to high-severity since 1984 (i.e., fire severity) was determined using fire-history data acquired from YNP (fire history data is available from the Integrated Resource Management Applications portal; severity data was acquired directly from the YNP fire atlas). For fire frequency we used the longest record available to calculate estimates: YNP fire perimeter records start in 1930. Most studies have described significant wildfire effects within ten years (Gresswell 1999), therefore, we used the proportion of each catchment burned ten years prior to sampling as our fire timing variable. Fire-severity values were calculated from Landsat Satellite Imagery, which first became available in 1984. Fire severity was estimated using normalized burn ratio values (NBR) calculated from Landsat 7 Enhanced Thematic Mapper Satellite

Imagery following Key and Benson (2006). Relative differences in normalized burn ratios (RdNBR) were calculated for each burned catchment. Breakpoints in RdNBR were determined for each pixel and assigned a value from zero to five. For the purpose of this study, values of 0 were considered to represent unburned, 1-2 to represent low-severity burned pixels, 3 to represent moderate-severity burned pixels, and 4-5 to represent high- severity burned pixels.

190

Ecosystem size

We used drainage area (km2) of each study reach as a coarse measure of ecosystem size and channel width as a finer-scale measure of local ecosystem size.

Drainage area was generated by delineating the catchment draining to each study reach using the watershed tool in ArcGIS. Channel width was determined by the average of the perpendicular width of the river channel at upstream, middle, and downstream locations within each sampling reach. The cross-sectional distance was estimated using the measure tool in ArcGIS from 1-m2 resolution aerial photographs taken in July 2014 by the US Department of Agriculture National Agriculture Imagery Program (NAIP).

Discharge

To measure hydrological disturbance, we estimated the magnitude of a reorganizing flood [i.e., 50% annual exceedance probability (50 AEP)] for each study reach along our transects using flow data from two gaged sites on the Merced River and regional flood frequency and magnitude equations for the Sierra Nevada (California

Region 3) developed by the US Geological Survey (Gotvald et al. 2012). For our gaged sites, we utilized the Pohono gage station for all study reaches downstream of the El

Capitan moraine and Happy Isles for all study reaches upstream of the El Capitan moraine. We used the following equation to determine the weighted flow estimate for ungaged sampling locations along our gradient: QP(u)w = [(2ΔA/Ag) + (1 -

2ΔA/Ag)(QP(g)w/QP(g)r)] * QP(u)r where QP(u)w is the weighted estimate of peak flow at the

191 ungaged sites in cubic feet per second; ΔA is the absolute value of the difference between the drainage areas of the gaged site and the ungaged site in square miles; Ag is the drainage area for the gaged site in square miles; and QP(u)r is the peak-flow estimate derived from the regional equations for the selected P-percent AEP at the gaged site in cubic feet per second. We then converted the data to cubic meters per second (m3 s-1).

Productivity

Although other researchers have used direct measurements of in situ primary productivity [e.g., chlorophyll α (Tonkin et al. 2013)], we did not make any direct measurements of benthic algal productivity in the field. Instead, we used Normalized

Difference Vegetation Index (NDVI) of sampling reaches as an estimate of potential in- stream primary productivity as degree of canopy openness has been linked to in-stream primary and secondary productivity (Wilzbach et al. 2005, Cooper et al. 2014). NDVI values are related to coverage by green vegetation, thus we assumed low NDVI values to signify a high-degree of canopy openness and therefore greater in-stream primary productivity. We first converted NAIP aerial photographs taken in July 2014 and obtained from the US Geological Survey (http://earthexplorer.usgs.gov/) to NDVI using the image analysis window in ArcGIS where NDVI = [(nir-red)/(red+nir)]. We selected the band arithmetic function in order to return NDVI values between -1 and 1 where more positive values indicate more cover by green vegetation and negative values indicate open water. We hand-delineated sampling reaches as 10x bankfull width (see ecosystem size section for channel width methods) using the original NAIP photographs,

192 then selected by mask the portion of the NDVI raster within our sampling reach. We used the mean NDVI value of each reach for analysis.

Statistical analysis

We used principle component analysis (PCA) to generate a fire axis from our fire frequency, severity, and timing variables (Table 9). All three variables represented strong, positive loadings on the 1st axis and this axis captured 98.22% of the total variance in the fire data set.

To examine our a priori predictions of the relative explanatory power of ecosystem size, disturbance, and productivity for tetragnathid spider and benthic macroinvertebrate predator trophic position and reliance on aquatically-derived energy, we first used an information-theoretic model selection approach based on Akaike’s information criterion (AIC) (Burnham and Anderson 2004). We used a correlation analysis to test for highly correlated (r > 0.8; Allen and Vaughn 2010, De Backer et al.

2010) independent variables and avoided using any highly-correlated variables in the same model (Burnham and Anderson 2004). Competing models were selected per response variable by linear regression. For each competing model, Akaike Information

Criterion adjusted for small sample size (AICc) was calculated. Models with ΔAICc ≤ 4 were retained as the most highly supported models (Burnham and Anderson 2004).

Akaike weights (ωi) were used to determine the relative support a model received among all of the candidate models in the set. The null model (i.e., intercept only) was also included in each of the sets of competing models for comparative purposes.

193

We then constructed path models (one candidate model per response variable) informed by a combination of predicted a priori relationships among variables, results of model- selection, and ecological plausibility. We used path analysis because of its suitability in representing direct and indirect relationships between hierarchically-structured, intercorrelated environmental variables (Shipley 1997, Shipley 2002). Maximum likelihood was used to estimate path coefficients and evaluate model fit and a χ2 test to determine overall model significance (i.e., when the χ2 p-value is > 0.10 the model is significantly different from the full, saturated model). We used several indices to judge model fit, including the comparative fit index (CFI), goodness-of-fit index (GFI), and root mean square error of approximation (RMSEA) (Hoyle 1995, Schumacker and

Lomax 2004). We performed regression analysis and generated AICc estimates using

JMP 11.0 (SAS Institute Inc., Cary, North Carolina, USA). We also explored potential non-linear relationships between trophic responses and fire by piecewise regression using the R package 'segmented' (Muggeo 2015). In a piecewise regression model, the relationship between the dependent and explanatory variable depends on whether the explanatory variable is above or below a certain value or 'breakpoint.' Model coefficients and breakpoints were estimated simultaneously using an iterative approach (Muggeo

2003). We compared linear and piecewise linear models using Akaike’s information criterion (AIC) and considered models with the lowest AIC value to be the best-fit

(Johnson and Omland 2004, Sasaki et al al. 2008). Path analysis was performed in

AMOS 22, an extension of SPSS (IBM SPSS Statistics, Armonk, NY) (Arbuckle 2013).

Because of the potential influence of underlying, unmeasured environmental variables

194 linked with the spatial distribution of our study reaches, we also tested for potential spatial autocorrelation (Moran’s I) of trophic position and reliance on aquatically-derived energy in ArcGIS 10.1.

Results

Study reaches along our gradient represented a range of wildfire effects on the landscape, from unburned to 11.3% burned two or more times since 1934 (frequency),

9.8% burned in the decade prior to sampling (2003-2012; time), and 7.3% burned with moderate-to-high severity since 1984 (dNBR ≥ 2; severity) (Table 10). Drainage area ranged from 0.44 (Red Peak Pass) to 114.76 (NPS Warehouse) km2 and channel width ranged from 2.92 (Red Peak Pass) to 57.97 (Bridalveil Creek confluence) m. Magnitude of a 50 AEP flood ranged from 0.17 m3 s-1 near the headwaters to 98.74 m3 s-1 near the park boundary, and NDVI values ranged from -0.50 to 0.11 where more positive values are associated with greater canopy cover over the channel (Table 10).

We collected tetragnathid spiders at all 31 study reaches and predatory benthic macroinvertebrates at 24 reaches. Predatory benthic macroinvertebrates primarily belonged to the order Plecoptera, however at a few study reaches we collected macroinvertebrates belonging to the orders Megaloptera (family: Corydalidae) and

Odonata (family Anisoptera). Of the Plecoptera, three families were identified:

Chloroperlidae, Perlidae, and Perlodidae.

Mean δ13C of tetragnathid spiders was -23.28‰ ± 2.36‰ (SD) and -23.82‰ ±

2.75‰ (SD) for predatory benthic macroinvertebrates. Mean δ15N was 3.26‰ ± 1.56‰

195 for spiders and 1.29‰ ± 1.70‰ for predatory benthic macroinvertebrates. Basal sources were sufficiently distinct for use in our mixing models (Post 2002a): mean δ13C was -

20.35‰ ± 5.50‰ for epilithic algae and -27.22‰ ± 1.25‰ for detritus. Mean δ15N for epilithic algae was -1.37‰ ± 2.78‰ and -2.22‰ ± 1.64‰ for detritus.

Reliance on aquatically-derived energy (i.e., from epilithic algae) ranged from

0.42-0.90 for spiders and 0.35-0.79 for predatory benthic macroinvertebrates, and the mean was slightly greater for spiders than for predatory benthic macroinvertebrates

[0.51% ± 0.12% (SD) and 0.48 ± 0.10 (SD), respectively; Table 11]. For both groups, reliance on aquatically-derived energy was relatively constant for roughly ¾ of the upstream end of the drainage gradient (0.30-0.59), then dramatically increased in the downstream ¼ of the gradient (0.60-0.90) (Figure 20).

Mean trophic position of spiders was 2.52 and ranged from 1.75 to 3.99. Mean trophic position of predatory benthic macroinvertebrates was 1.95 and ranged from 1.02 to 3.02 (Table 11). Tetragnathid spider trophic position was slightly lower in the upstream-most study reaches, highest in the downstream-most study reaches, and moderate in the middle sampling locations (Figure 21). This pattern was echoed by predatory benthic macroinvertebrate trophic position, however trophic position of predatory benthic macroinvertebrates was consistently lower than spider trophic position, especially at upstream study reaches (Figure 21). Moran’s I indicated no spatial autocorrelation for either of the Tetragnathidae response variables (i.e., reliance on aquatically-derived energy or trophic position; Moran’s I: p ≤ 0.001; Table 11), however both of these metrics were spatially autocorrelated for predatory benthic

196 macroinvertebrates, indicating that the spatial distributions of predatory benthic macroinvertebrate trophic position and reliance on aquatically-derived energy were more spatially clustered than if underlying spatial processes were random.

Model selection

Flood magnitude and the fire axis were highly correlated with drainage area (r =

0.96 and 0.93 for flood magnitude and fire axis, respectively), therefore both of these variables were not entereted into the same model. Model-selection results indicated that flood magnitude, channel width, and NDVI were important predictors of reliance on aquatically-derived energy by tetragnathid spiders (Table 12). We found support for seven models explaining variation in predatory benthic invertebrate reliance on aquatically-derived energy with flood magnitude again emerging as a predictor in the top models of the set. Drainage area, channel width, NDVI, and the fire axis also received support (Table 12). Likewise, flood magnitude, drainage area, and NDVI (but not fire) were salient variables for trophic position of riparian spiders. Channel width was the strongest predictor variable for trophic position of benthic macroinvertebrate predators, although the fire axis, drainage area, flood magnitude, and NDVI also received support

(Table 12).

We developed one plausible path model for each response variable based on our a priori hypotheses and the results of model selection, with measures of ecosystem size included as the independent variables in all models (Figure 22). Drainage area exhibited a direct, negative relationship with trophic position of both groups of invertebrate predators

197

(Figure 22). Whereas channel width also had a weak negative, direct effect on spider trophic position, it had a strong positive, direct effect on benthic macroinvertebrate predator trophic position (Figure 22). Flood magnitude emerged as a strong indirect pathway for three of the four models: reliance on aquatically-derived energy for tetragnathid spiders (Figure 22a), reliance on aquatically-derived energy for benthic macroinvertebrate predators (Figure 22b), and trophic position of tetragnathid spiders

(Figure 22c). In all three models, flood magnitude had a positive effect with greater standardized regression weights than either measure of ecosystem size indicating it as a mediator of ecosystem size for determining trophic dynamics of invertebrate consumers.

NDVI was a key mediator for both reliance on aquatically-derived energy for benthic invertebrate predators (weak positive effect; Figure 22b) and trophic position (negative effect) of benthic invertebrate predators (Figure 22d). The fire axis was only included as an indirect pathway for trophic position of benthic invertebrate predators with a weak negative effect (Figure 22d).

Piecewise linear relationships

The distribution of both trophic response variables and fire pointed to possible non-linear patterns relative to drainage area (Figure 23). For example, at the confluence of Illilouette Creek and the main stem of the Merced River, drainage area increased from

33.1 to 52.2 km2; a 166% increase. However, at this same location, proportion of the catchment burned greater than two times since 1930 increased from 0.7 to 10.0%, the proportion of the catchment burned one decade prior to sampling increased from 1.2 to

198

5.8% and the proportion of the catchment burned with moderate-to-high severity increased from 0.8 to 4.6%; therefore the proportion of the catchment affected by fire increased by 500-1400% (Figure 23).

Piecewise linear regression indicated statistical support for breakpoints in tetragnathid spider responses for both reliance on aquatically-derived energy and trophic position. For aquatically-derived energy, the first breakpoint along the fire axis ([1.92 ±

0.23 (SE)] corresponds with the confluence of the Illilouette drainage with the Merced

River (as described in the preceding paragraph). Fitting the segmented linear function to the relationship between fire and reliance on aquatically-derived energy by tetragnathid spiders increased the amount of variation explained from 55% (linear regression) to 82%

(Figure 24a) and revealed a positive relationship between the proportion of each catchment burned with frequent, recent, or severe fire and reliance on aquatically-derived energy by spiders following the detected breakpoint. A segmented linear fit of drainage area with reliance on aquatically-derived energy identified a breakpoint at 99.07 km2 ±

2.04 km2 (SE), revealing a sharp increase in the positive relationship between drainage area and reliance on aquatically-derived energy by tetragnathid spiders, and 84% of the variation (as opposed to 48% for a linear regression; Figure 24b).

We detected two breakpoints in the linear relationship between the fire axis and trophic position of tetragnathid spiders at -0.87 ± 0.26 (SE) and 3.62 ± 0.08 (SE), the latter of which was primarily driven by only two data points (Figure 24c). Fitting a segmented linear function to the relationship between the fire axis and trophic position of spiders increased the amount of variation explained from 42% (linear regression) to 75%,

199

In addition, at the first breakpoint the relationship between the fire axis and trophic position changed from positive to negative. A segmented linear fit of drainage area with trophic position increased the amount of variation explained from 43% (linear regression) to 70%, indicated support for a breakpoint at 104.70 km2 ± 2.55 (SE) km2, and similar to the second breakpoint detected in the linear relationship between the fire axis and TP, this breakpoint also appeared to be driven by only two data points (Figure 24d).

Discussion

Examination of trophic dynamics along a gradient of drainage area size from the headwaters of the Merced River to 4th- and 5th- order segments near the boundary of

Yosemite National Park in the central Sierra Nevada revealed significant variability in the relative reliance on aquatically-derived energy (i.e., derived from benthic algal pathways) and trophic position by/of invertebrate stream-riparian predators. Although ecosystem size (i.e., drainage area and channel width) was strongly related to both trophic measures, flood magnitude appeared to be a common indirect linkage between ecosystem size and trophic responses. Invertebrate consumer reliance on aquatic primary production and trophic position increased markedly from roughly 50 to 80-90% in stream reaches draining catchments > 100 km2, suggesting that trophic responses to fire may be non- linear (Holt 2004, Arim et al. 2007). With few exceptions, we did not find strong support for fire as an indicator of trophic dynamics in this system, but the maximum area affected by fire in our study was only 11.3% and fire scaled with drainage area, which may have limited our ability to both detect and partition out fire effects on the landscape.

200

Reliance on aquatically-derived energy

In general, food webs in mid-order streams (e.g., Merced River near the YNP boundary) tend to be relatively more reliant on autochthonous primary production than their upstream counterparts (Minshall 1978, Vannote et al. 1980, Rosi-Marshall and

Wallace 2002), therefore we expected a positive relationship between drainage area and reliance on aquatically-derived energy for both benthic macroinvertebrate predators as well as tetragnathid spiders. In addition, greater channel width should be positively related to reliance on in situ primary production as a function of increased penetration of solar radiation to stream benthos (Davies-Colley and Quinn 1998). Although model- selection results indicated a negative effect of drainage area and channel width on reliance on aquatically-derived energy by both invertebrate predator taxa, path models indicated a positive effect of ecosystem size via indirect effects of flood magnitude, fire, and productivity (NDVI). For example, although the direct effect of channel width was negative in the path model explaining reliance on aquatically-derived energy sources by predatory benthic macroinvertebrates, NDVI acted as a mediator whereby broader channels also had less canopy cover resulting in a net positive effect presumably through increased light penetration and increased autochthonous productivity (Figure 22b).

Floods are a primary driver of flows of organisms, nutrients, and energy from the stream to the riparian zone (e.g., Junk et al. 1989), and in our study although the direct effect of channel width was negative for models explaining reliance on aquatically-derived energy sources by benthic and riparian predators (Figure 22a and b), the inclusion of flood

201 magnitude as a mediator between channel width and our response variables resulted in a net positive relationship between channel width and reliance on aquatically-derived energy by invertebrate predators. Both direct and indirect effects of flood magnitude may be driving predatory invertebrate responses (see following paragraphs).

High-magnitude floods constitute a significant disturbance (Resh et al. 1988) that can have pronounced effects on both basal resources [i.e., benthic algae, fine particulate organic matter (FPOM), and coarse particulate organic matter (CPOM)] and primary consumers (i.e., benthic invertebrates that constitute a prey source for both groups of invertebrate predators considered in this study). High flows can remove benthic algae

(Robinson et al. 2004b) as well as flush out CPOM (Brookshire and Dwire 2003) and these effects can last years after a high-flow event (Robinson et al. 2003). Benthic macroinvertebrates are similarly transported downstream during high-flows (Robinson et al. 2004a) and recolonizing benthic invertebrates are dominated by grazers and generally follow the recovery trajectory of benthic algae. Further, benthic macoinvertebrate community composition in streams characterized by regular high-magnitude floods tend to retain this character compared to streams with more dampened hydrographs (Robinson et al. 2004a), indicating a lasting reliance on benthic algae as a food resource for primary consumers. Therefore, both aquatic and riparian consumers of benthic insects might be expected to reflect a stronger aquatically-derived energy pathway in systems influenced by high-magnitude flows.

Riparian tetragnathid spiders build their webs directly over or adjacent to streams and can selectively prey on adult aquatic insects as they emerge from the stream (Collier

202 et al. 2002); however, unlike benthic macroinvertebrate predators, tetragnathid spiders also prey on terrestrial insects. For example, Kato et al. (2003) found that tetragnathid spiders were seasonally dependent on aquatic prey and relied more heavily on terrestrial prey when aquatic subsidies from streams were low. Tetragnathid spider reliance on aquatically-derived energy has been found to be related to overhanging vegetation, canopy density, and the density and body size of emergent aquatic insects, suggesting that aquatically-derived energy is thus a function of both terrestrial and aquatic components of stream corridors (Tagwireyi and Sullivan 2015). Consequently, we expected that tetragnathid spiders would be more reliant on aquatically-derived energy where they have greater access to aquatic prey and stream habitat. Along with flood magnitude, channel width garnered the greatest weight of evidence as a predictor of tetragnathid spider reliance on aquatically-derived energy (Table 12) and we infer that emergent aquatic insects may be less accessible to tetragnathid spiders where channel width is wide and the perimeter (bank length) to area (bank length x channel width) ratio is low, however high-magnitude floods may change the nature of the rlationship through inundation of riparian zones as supported by our hypothesized path model (Figure 22a).

Trophic position

Trophic position of top invertebrate consumers in this study was heavily influenced by ecosystem size at both the catchment (i.e., drainage area) and local (i.e., channel width) scales with hydrological disturbance implicated as an important mechanism driving observed patterns (Figure 22c). Previous studies have observed a

203 negative relationship between hydrologic disturbance and FCL (Parker and Huryn 2006,

McHugh et al. 2010, Sabo et al. 2010), or no relationship at all (Thompson and

Townsend 1999), however we found a positive relationship between flood magnitude and trophic position of tetragnathid spiders (although note that one supported model for benthic macroinvertebrate predators indicated a negative relationship, Table 12). Sabo et al. (2010) found that flow variability was a key mechanism driving links between ecosystem size and FCL in streams, and that more variable flows were associated with shorter FCL. This result was related to both stochastic and predictable flow variation.

Other studies that have directly measured hydrologic shifts in benthic substrate (e.g.,

Thompson and Townsend 1999), or estimated flow variability through a combination of coefficients of variation (e.g., McHugh et al. 2010, Sabo et al. 2010), have found a negative relationship between flood magnitude and FCL. However, predictable floods can increase FCL in streams. For example, annual scouring flows in the California’s coastal Eel River temporarily extirpate armored caddisflies (Power et al. 2008), which can result in a release of epilithic algae from grazing pressure and a bottom-up increase in density of benthic invertebrate prey, abundance of insectivorous fish, and an increase in overall FCL (Marks et al. 2000). Similar disturbance-related community dynamics may be important in our system where trophic position of top invertebrate consumers was linked to increased flood magnitude estimates, which reflect flood events of a certain size at predictable time intervals (Gotvald et al. 2012).

Riparian consumers have not traditionally been considered within FCL contexts in fluvial systems, but represent integral components of stream-riparian food webs

204

(reviewed in Baxter et al. 2005, Sullivan and Rodewald 2012). Tagwireyi and Sullivan

(2015) estimated a trophic position of 3.12 for tetragnathid spiders in a 5th-6th- order river of Ohio (vs. 2.52 in our study) and report that trophic position was related to canopy density, shoreline geometry, overhanging vegetation coverage, and emergent aquatic insect density. In our Yosemite study, tetragnathid spider trophic position was influenced by ecosystem size at both the catchment and local scales with direct, negative relationships with both drainage area and channel width (Table 12, Figure 22c). As in other studies (e.g., Sabo et al. 2010), hydrologic variability acted as a strong mediator

(although the directionality of the relationships differed). In addition, NDVI (used as a proxy for productivity) was supported in two of five models explaining spider trophic position and had a moderate negative effect; that is, reaches with more open water supported spiders feeding at a higher trophic position. Taken together these results indicate that emergent aquatic insect density may be the most important factor affecting spider trophic position in this system, as trophic position was higher in larger, more open stream reach sections that should be more productive (i.e., facilitating a greater density of emergent aquatic invertebrates) (Kato et al. 2003, Malison and Baxter 2010b), although we did not measure this directly.

Wildfire effects

Fire was not as strongly supported a predictor of trophic dynamics as ecosystem size, flood magnitude, or productivity in our study. One exception was as a predictor of predatory benthic macroinvertebrate trophic position (Figure 22d). In this path model,

205 fire exerted a negative effect on benthic macroinvertebrate predator trophic position as a mediator between drainage area and trophic position. In the same path model, increased channel width was directly related to increased trophic position of benthic macroinverterbate predators with opening of the canopy (more negative NDVI values) included as a significant indirect effect. Therefore, whereas ecosystem size and productivity were associated with increased trophic position of predatory benthic macroinvertebrates, which aligns with observations by others (Townsend et al. 1998,

McHugh et al. 2010), fire had a notable negative effect. Fire can lead to decreases in the taxonomic richness of benthic invertebrate communities (Minshall et al. 2003, Vieira et al. 2004), and decreased richness is associated with lower FCL (Cohen and Newman

1991), presumably due to removal of mid-level predators and an increase in omnivory

(Yodzis 1984), which may be one mechanism explaining the negative effect of fire in this example.

Integrating non-linear effects of wildfire might increase our ability to detect significant effects of fire on trophic dynamics in streams. We predicted that patterns in invertebrate trophic responses would be discontinuous along gradients in the relative proportion of the catchment affected by wildfire. In fact, post-hoc exploration of non- linear responses to fire for tetragnathid spider reliance on aquatically-derived energy

(Figure 24a) and trophic position (Figure 24c) yielded models with greater support than linear models. A breakpoint in the piecewise linear relationship between the fire axis and reliance on aquatically-derived energy by spiders was associated with a dramatic increase in the proportion of each catchment burned with severe, frequent, or recent fire (Figure

206

23 and Figure 24a), which aligned with the confluence of the Illilouette catchment with the Merced River. At confluences draining catchments heavily influenced by fire, increased supply of sediments and debris may occur (Benda et al. 2003, Wondzell and

King 2003, Shakesby and Doerr 2006) and be associated with increased export of aquatic prey (Harris et al. In press), which may result in greater availability of emergent aquatic insect prey to spiders (i.e., Malison and Baxter 2010b), and therefore greater reliance on aquatically-derived energy if emergent aquatic insects are feeding on algae. Although whether this is the case in our study is not clear from our results.

A breakpoint in the linear relationship between the fire axis and trophic position of tetragnthid spiders was statistically supported at -0.87 where we observed a sharp spike in the proportion of each catchment burned with severe, frequent, or recent wildfire

(Figure 23 and Figure 24c). Beyond this breakpoint the relationship between the fire axis and trophic position shifted from positive to negative (Figure 24c). Likewise, fire was also inversely related to trophic position of predatory benthic macroinvertebrates (Table

12, Figure 22d). Others have observed that wildfire can decrease taxonomic richness of benthic invertebrate communities (Minshall et al. 2003, Vieira et al. 2004) and decreased taxonomic richness is linked to shorter FCL (Cohen and Newman 1991), potentially caused by increased omnivory and removal of mid-level predators (Wang et al. 2013).

Although, we did not measure taxonomic richness in this study, this may be one explanation for observed negative relationships between wildfire and trophic position of tetragnathid spiders and predatory benthic macroinvertebrates

207

All of our fire variables were highly correlated with drainage area in this study, likely due to increased fire activity at lower elevations (and thus larger drainage areas).

An experimental design that partitions out this correlation (i.e., compares catchments at similar elevations, but varying by drainage area and proportion influenced by fire) would go a long way toward identifying fire-specific effects.

Conclusions

Ecosystem size at both the landscape and local scale was most predictive of trophic dynamics in our study with flood magnitude indicated as an important mechanism. Disturbance in the form of floods are a major driver of stream ecosystem structure and function (Resh et al. 1988) and flow variability has been implicated as a key mechanism facilitating ecosystem size effects on FCL (Sabo et al. 2010). Further, hydrologic regimes can be the only significant control on FCL in systems that have intermittent precipitation patterns (Warfe et al. 2013), and thus the predictive nature of high magnitude floods in our Mediterranean-climate system is not surprising. Flow regimes also interact strongly with wildfire to generate shifts in stream geomorphology

(Miller et al. 2003, Shakesby and Doerr 2006, May 2007), primary productivity (Davis et al. 2013), and benthic invertebrate abundance and community composition (Arckle et al.

2010, Verkaik et al. 2013) in streams. However, largely constrasting our expectations that fire would represent an additional important disturbance agent governing stream-riparian invertebrate trophic dynamics, fire was not highly supported. However, strong non-linear models provided preliminary evidence that invertebrate trophic responses may be related

208 to thresholds in both fire and ecosystem size. Therefore, further inquiry into the dynamics between fire, floods, and stream-riparian food webs may be needed to reveal significant effects with implications for ecological theory related to environmental control on trophic dynamics. Understanding the interactions among these drivers will also be important for ecosystem management of stream-riparian systems in fire-prone regions where dynamic fire regimes that operate over time and space may be important in maintaining ecosystem integrity and biodiversity of linked stream-riparian ecosystems (Jackson et al. In press).

Acknowledgements

Funding was provided by NSF DEB-1401480 awarded to SMPS and BKJ, Bureau of

Land Management (14-3-01-37) award to SMPS, and The Ohio State University. In

Yosemite National Park we received help from Dr. G. Smith, S. Stock, J. Roche, and K.

Van Wagtendonk. We appreciate field and laboratory assistance received from D. Groff,

M. Hickson, Dr. K. Hossler, P. Koubek, M. Ledford, L. Meyer, K. Zhao, and D. Vent.

209

Literature Cited

Alberts, J. M., S. M. P. Sullivan, and A. Kautza. 2013. Riparian swallows as integrators of landscape change in a multiuse river system: implications for aquatic-to- terrestrial transfers of contaminants. Science of the Total Environment 463:42-50. Allan, J. D., M. S. Wipfli, J. P. Caouette, A. Prussian, and J. Rogers. 2003. Influence of streamside vegetation on inputs of terrestrial invertebrates to salmonid food webs. Canadian Journal of Fisheries and Aquatic Sciences 60:309-320. Allen, D. C. and C. C. Vaughn. 2010. Complex hydraulic and substrate variables limit freshwater mussel species richness and abundance. Journal of the North American Benthological Society 29:383-394. Arbuckle, J. L. 2013. Amos (Version 22.0) SPSS, Chicago. Arim, M., P. A. Marquet, and F. M. Jaksic. 2007. On the relationship between productivity and food chain length at different ecological levels. American Naturalist 169:62-72. Arkle, R. S., D. S. Pilliod, and K. Strickler. 2010. Fire, flow and dynamic equilibrium in stream macroinvertebrate communities. Freshwater Biology 55:299-314. Baxter, C. V., K. D. Fausch, M. Murikami, and P. L. Chapman. 2004. Fish invasion restructures stream and forest food webs by interrupting reciprocal prey subsidies. Ecology 85:2656-2663. Baxter, C. V., K. D. Fausch, and W. C. Saunders. 2005. Tangled webs: reciprocal flows of invertebrate prey link stream and riparian zones. Freshwater Biology 50:201- 220. Bisson, P.A., B.E., Rieman, C., Luce, P.F., Hessburg, D.C., Lee, J.L., Kersner, G.H., Reeves, and R.E., Gresswell. 2003. Fire and aquatic ecosystems of the Western USA: current knowledge and key questions. Forest Ecology and Management 178, 213–229. Bonada, N. and V. H. Resh. 2013. Mediterranean-climate streams and rivers: geographically separate but ecologically comparable freshwater systems. Hydrobiologia 719:1-29. Brookshire, E. N. J. and K. A. Dwire. 2003. Controls on patterns of coarse organic particle retention in headwater streams. Journal of the North American Benthological Society 22:17-34. Burnham, K. P. and D. R. Anderson. 2004. Multimodel inference - understanding AIC and BIC in model selection. Sociological Methods & Research 33:261-304. Carpenter, S. R., J. J. Cole, M. L. Pace, M. Van de Bogert, D. L. Bade, D. Bastviken, C. M. Gille, J. R. Hodgson, J. F. Kitchell, and E. S. Kritzberg. 2005. Ecosystem subsidies: terrestrial support of aqautic food webs from 13C addition to contrasting lakes. 86 10:2737-2750 210

Cohen, J. E. and C. M. Newman. 1991. Community area and food chain length: theoretical predictions. American Naturalist 138:1542-1554. Collier, K. J., J. S. Bury, and M. Hibbs. 2002. A stable isotope study of linkages between stream and terrestrial food webs through spider predation. Freshwater Biology 47:1651-1659. Cooper, S. D., H. M. Page, S. W. Wiseman, K. Klose, D. Bennet, T. Even, S. Sadro, C. E. Nelson, and T. L. Dudley. 2014. Physicochemical and biological responses of streams to wildfire severity in riparian zones. Freshwater Biology. Davies-Colley, R. J. and J. M. Quinn. 1998. Stream lighting in five regions of North Island, New Zealand: control by channel size and riparian vegetation. New Zealand Journal of Marine and Freshwater Research 32:591-605. Davis, J. M., C. V. Baxter, E. J. Rosi-Marshall, J. L. Pierce, and B. T. Crosby. 2013. Anticipating stream ecosystem responses to climate change: toward predictions that incorporate effects via land-water linkages. Ecosystems 16:909-922. De Backer, A., C. Van Colen, M. Vincx, and S. Degraer. 2010. Complex hydraulic and substrate variables limit freshwater mussel species richness and abundance. Journal of the North American Benthological Society 29:383-384. DeFlorio, M. J., D. W. Pierce, D. R. Cayan, and A. J. Miller. 2013. Western US extreme precipitation events and their relation to ENSO and PDO in CCSM4. Journal of Climate 26:4231-4243. Dunham, J. B., A. E. Rosenberger, C. H. Luce, and B. E. Rieman. 2007. Influences of wildfire and channel reorganization on spatial and temporal variation in stream temperature and distribution of fish and amphibians. Ecosystems 10:335-346. Finlay, J. C. 2001. Stable-carbon-isotope ratios of river biota, implications for energy flow in lotic food webs. Ecology 82:1052-1064. Gotvald, A. J., N. A. Barth, A. G. Veilleux, and C. Parrett. 2012. Methods for determining magnitude anf frequency of floods in California, based on data through water year 2006. available online only at http://pubs.usgs.gov/sir/2012/5113/. Gresswell, R. E. 1999. Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193-221. Harris, H. E., C. V. Baxter, and J. M. Davis. In press. Debris flows mediate effects of wildfire on magnitude and composition of tributary subsidies to main-stem habitats. Freshwater Science. Hicks, J. H., M. S. Wipfli, D. W. Lang, and M. E. Lang. 2005. Marine-derived nitrogen and carbon in freshwater-riparian food webs of the Copper River Delta, southcentral Alaska. Oecologia 144:558-569.

211

Holt, R. D. 2004. Implications of system openess for local community structure and ecosystem function.in G. A. Polis, M. E. Power, and G. R. Huxel, editors. Food webs at the landscape level. The University of Chicago Press, Chicago, IL, U.S.A. Hoyle, R. H. 1995. Structural equation modeling: concepts, issues, and applications. Sage Publications Inc., Thousand Oaks, CA. Jackson, B. K. and S. M. P. Sullivan. In press. Responses of riparian tetragnathid spiders to wildfire in forested ecosystems of the California Mediterranean climate region, USA. Freshwater Science. Jackson, B. K., S. M. P. Sullivan, C. V. Baxter, and R. L. Malison. In press. Stream- riparian ecosystems.in D. A. Dellasala and C. T. Hanson, editors. Benefits of high-severity fire for biodiversity and conservation. Elsevier. Jackson, B. K., S. M. P. Sullivan, and R. L. Malison. 2012. Wildfire severity mediates fluxes of plant material and terrestrial invertebrates to mountain streams. Forest Ecology and Management 278:27-34. Johnson, J.B. and K.S. Ormland. 2004. Model selection in ecology and evolution. Trends in Ecology and Evolution 19:101-108. Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood pulse concept in river- floodplain systems. Pages 110-127 International large river symposium. Kato, C., T. Iwata, S. Nakano, and D. Kishi. 2003. Dynamics of aquatic insect flux affects distribution of riparian web-building spiders. Oikos 103:113-120. Key, C.H. and N.C. Benson. 2006. Landscape assessment: ground measure of severity, the Composite Burn Index; and remote sensing of severity, the Normalized Burn Ratio. In D.C. Lutes; R.E. Keane; J.F. Caratti; C.H. Key; N.C. Benson; S. Sutherland; and L.J. Gangi. 2006. FIREMON: Fire effects monitoring and inventory system. USDA Forest Service, Rocky Mountain Research Station, Ogden, UT. Gen. Tech. Rep. RMRS-GTR-164-CD: LA 1-51. Malison, R. L. and C. V. Baxter. 2010a. Effects of wildfire of varying severity on benthic stream insect assemblages and emergence. Journal of the North American Benthological Society 29:1324-1338. Malison, R. L. and C. V. Baxter. 2010b. The fire pulse: wildfire stimulates flux of aquatic prey to terrestrial habitats driving increases in riparian consumers. Canadian Journal of Fisheries and Aquatic Sciences 67:570-579. Marks, J. C., M. E. Power, and M. S. Parker. 2000. Flood disturbance, algal productivity, and interannual variation in food chain length. Oikos 90:20-27. May, C. 2007. Sediment and wood routing in steep headwater streams: An overview of geomorphic processes and their topographic signatures. Forest Science 53:119- 130. McHugh, P. A., A. R. McIntosh, and P. G. Jellyman. 2010. Dual influences of ecosystem size and disturbance on food chain length in streams. Ecology Letters 13:881-890. 212

Meyer, L. A. and S. M. P. Sullivan. 2013. Bright lights, big city: influences of ecological light pollution on reciprocal stream-riparian invertebrate fluxes. Ecological Applications 23:1322-1330. Miller, D., C. Luce, and L. Benda. 2003. Time, space, and episodicity of physical disturbance in streams. Forest Ecology and Management 178:121-140. Minshall, G. W. 1978. Autotrophy in stream ecosystems. Bioscience 28:767-770. Minshall, G. W. 2003. Responses of stream benthic macroinvertebrates to fire. Forest Ecology and Management 178:155-161. Minshall, G. W., C. T. Robinson, D. E. Lawrence, D. A. Andrews, and J. T. Brock. 2001. Benthic macroinvertebrate assemblages in five central Idaho (USA) streams over a 10-year period following disturbance by wildfire. International Journal of Wildland Fire 10:201-213. Muggeo, V M R. 2015. Segmented: Regression models with breakpoints/changepoints estimation. R package version 0.5-1.1. Available at: http://cran.r- project.org/web/packages/segmented/index.html. Muggeo, V.M.R. 2003. Estimating regression models with unknown break-points. Statistics in Medicine, 22: 3055-3071. Nakano, S., H. Miyasaka, and N. Kuhara. 1999. Terrestrial-aquatic linkages: riparian arthropod inputs alter trophic cascades in a stream food web. Ecology 80:2435- 2441. Paetzold, A. C., C. J. Schubert, and K. Tockner. 2005. Aquatic-terrestrial linkages along a braided river: riparian arthropods feeding on aquatic insects. Ecosystems 8:748- 758. Parker, S. M. and A. D. Huryn. 2006. Food web structure and function in two arctic streams with contrasting disturbance regimes. Freshwater Biology 51:1249-1263. Parnell, A., R. Inger, S. Bearhop, and A. Jackson. 2010. Coping with too much variation. . PLoS ONE 5:e9672. Parnell, A. and A. Jackson. 2013. Stable Isotope Analysis in R. http://cran.r- project.org/web/pakage/siar/. Piccolo, J. J. and M. S. Wipfli. 2002. Does red alder (Alnus rubra) in upland riparian forests elevate macroinvertebrate and detritus export from headwater streams to downstream habitats in southeastern Alaska? Canadian Journal of Fisheries and Aquatic Sciences 59:503-513. Pimm, S. L. and J. H. Lawton. 1977. Number of trophic levels in ecological communities. Nature 268:329-331. Post, D. M. 2002a. The long and short of food-chain length. Trends in Ecology & Evolution 17:269-278.

213

Post, D. M. 2002b. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83:703-718. Power, M. E. 2002. River-watershed exchange: effects of riverine subsidies on riparian lizards and their terrestrial prey. Ecology 83. Power, M. E., M. S. Parker, and W. E. Dietrich. 2008. Seasonal reassembly of a river food web: floods, droughts, and impacts of fish. Ecological Monographs 78:263- 282. Resh, V. H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W. Li, G. W. Minshall, S. R. Reice, A. L. Sheldon, J. B. Wallace, and R. C. Wissmar. 1988. The role of disturbance in stream ecology. Journal of the North American Benthological Society 7:433-455. Robinson, C. T., S. R. Rushforth, and G. W. Minshall. 1994. Diatom asssemblages of streams influenced by wildfire. Journal of Phycology 30:209-216. Robinson, C. T., K. Tockner, and P. Burgherr. 2004a. Drift benthos relationships in the seasonal colonization dynamics of alpine streams. Archiv Fur Hydrobiologie 160:447-470. Robinson, C. T., U. Uehlinger, and M. T. Monaghan. 2003. Effects of a multi-year experimental flood regime on macroinvertebrates downstream of a reservoir. Aquatic Sciences 65:210-222. Rosi-Marshall, E. J. and J. B. Wallace. 2002. Invertebrate food webs along a stream resources gradient. Freshwater Biology 47:129-141. Royer, T. V. and G. W. Minshall. 1997. Temperature patterns in small streams following wildfire. Archiv Fur Hydrobiologie 140:237-242. Rugenski, A. T. and G. W. Minshall. 2014. Climate-moderated responses to wildfire by macroinvertebrates and basal food resources in montane wilderness streams. Ecosphere 5. Sabo, J. L., J. C. Finlay, T. Kennedy, and D. M. Post. 2010. The role of discharge variation in scaling of drainage area and food chain length in rivers. Science 330:965-967. Sabo, J. L., J. C. Finlay, and D. M. Post. 2009. Food chains in freshwaters. Pages 187- 220 in R. S. Ostfeld and W. H. Schlesinger, editors. Year in Ecology and Conservation Biology 2009. Sasaki, T., T. Okayasu, J. Undarmaa, and K. Takeuchi. 2008. Changes in vegetation along a grazing gradient in Mongolian rangelands. Journal of Ecology 96: 145- 154 Schumacker, R. E. and R. G. Lomax. 2004. A beginner's guide to structural equation modeling. Lawrence Erlbaum and Associates Inc., Mahwah, NJ.

214

Shakesby, R. A. and S. H. Doerr. 2006. Wildfire as a hydrological and geomorphological agent. Earth-Science Reviews 74:269-307. Shipley, B. 1997. Exploratory path analysis with applications in ecology and evolution. American Naturalist 149:1113-1138. Shipley, B. 2002. Cause and correlation in biology: A user's guide to path analysis, structural equations and casual inference. Spencer, C. N., K. Odeny-Gabel, and F. R. Hauer. 2003. Wildfire effects on stream food webs and nutrient dynamics in Glacier National Park, USA. Forest Ecology and Management 178:141-152. Sullivan, S. M. P. and A. D. Rodewald. 2012. In a state of flux: the energetic pathways that move contaminants from aquatic to terrestrial environments. Environmental Toxicology and Chemistry 31:1175-1183. Tagwireyi, T. and S. M. P. Sullivan. 2015. Distribution and trophic dynamics of riparian tetregnathid spiders in a large river system. Marine and Freshwater Research. Thompson, R. M. and C. R. Townsend. 1999. The effect of seasonal variation on the community structure and food-web attributes of two streams: implications for food-web science. Oikos 87:75-88. Tonkin, J. D., R. G. Death, and K. J. Collier. 2013. Do productivity and disturbance interact to modulate macroinvertebrate diversity in streams? Hydrobiologia 701:159-172. Townsend, C. R., R. M. Thompson, A. R. McIntosh, C. Kilroy, E. Edwards, and M. R. Scarsbrook. 1998. Disturbance, resource supply, and food-web architecture in streams. Ecology Letters 1:200-209. Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Canadian Journal of Fishery and Aquatic Science 37:130-137. Verkaik, I., M. Vila-Escale, M. Rieradevall, and N. Prat. 2013. Seasonal drought plays a stronger role than wildfire in shaping macroinvertebrate communities of Mediterranean streams. International Review of Hydrobiology 98:271-283. Vieira, N. K., W. H. Clements, L. S. Guevara, and B. F. Jacobs. 2004. Resistance and resilience of stream insect communities to repeated hydrologic disturbances after a wildfire. Freshwater Biology 49:1243-1259. Voshell, J. R. 2002. A guide to common freshwater invertebrates of North America. The McDonald and Woodward Publishing Company, Blacksburg, Virginia. Wan, S., D. Hui, and Y. Luo. 2001. Fire effects on nitrogen pools and dynamics in terrestrial ecosystems: a meta-analysis. Ecological Application 11: 1349-1365. Warfe, D. M., T. D. Jardine, N. E. Pettit, S. K. Hamilton, B. J. Pusey, S. E. Bunn, and P. M. Davies. 2013. Productivity, disturbance and ecosystem size have no influence

215

on food chain length in seasonally connected rivers. PLoS ONE 10.1371/journal.pone.0066240. Wilzbach, M. A., B. C. Harvey, J. L. White, and R. J. Nakamoto. 2005. Effects of riparian canopy opening and salmon carcass addition on the abundance and growth of resident salmonids. Canadian Journal of Fisheries and Aquatic Sciences 62:58-67. Wondzell, S. M. and J. G. King. 2003. Postfire erosional processes in the Pacific Northwest and Rocky Mountain regions. Forest Ecology and Management 178:75-87. Yodzis, P. 1984. Energy flow and the vertical structure of real ecosystems. Oecologia 65:86-88.

216

Principle Component “Fire Axis” Independent variable loading r2 Fire frequency 0.99 2.89 Fire timing 0.99 2.89 Fire severity 0.99 2.89

Eigenvalue 2.95 Percent variance explained 98.22

217

Table 10. Minimum, maximum, mean ( ), and standard deviation (SD) of predictor variables relating to ecosystem size (drainage

area and channel width), productivity (NDVI), and disturbance (flood magnitude and fire frequency, timing, and severity) for study

reaches along a gradient of drainage area size extending from the headwaters of the Merced River and South Fork of the Merced

River to the 4th-order mainstem of the South Fork of the Merced River and 5th-order mainstem of the Merced River.

Predictor variables Minimum Maximum SD Ecosystem size Drainage area (km2) 0.44 114.76 43.05 38.95

Channel width (m) 2.92 57.97 25.04 14.54 Productivity

NDVI -0.50 0.11 -0.27 0.14 Disturbance

Proportion of catchment burned more than two times after 1930 (%) 0.00 11.30 3.27 4.25 Proportion of catchment burned ten or fewer years prior to sampling (%) 0.00 9.79 2.62 3.18 Proportion of catchment burned with moderate-to-high severity after 1984 (%) 0.00 7.25 1.81 2.29 Magnitude of a 50 AEP flood (m3 s-1) 0.17 98.74 25.24 29.26

b 218

Table 11. Mimimum, maximum, mean ( ), and standard deviation (SD) of response variables including reliance on aquatically-

derived energy (i.e., proportion of nutritional subsidies derived from benthic algal pathways) and trophic position of tetragnathid

spiders and benthic macroinvertebrate predators along a gradient of drainage area size extending from the headwaters of the

Merced River and South Fork of the Merced River to the 4th-order mainstem of the South Fork of the Merced River and 5th-order

mainstem of the Merced River. In addition, z and p values from spatial autocorrelation analysis using Moran’s I are presented for

each response variable.

9 Response variables Minimum Maximum SD Moran's I z p

Reliance on aquatically-derived energy Tetragnathid spiders 0.42 0.90 0.51 0.12 0.49 3.66 < 0.001 Predatory benthic macroinvertebrates 0.35 0.79 0.48 0.11 0.03 0.44 0.657 Trophic position Tetragnathid spiders 1.75 3.99 2.52 0.43 0.46 3.44 0.001 Predatory benthic macroinvertebrates 1.02 3.02 1.95 0.52 0.04 0.46 0.642

b 219

Table 12. Retained regression models (ΔAICc ≤ 4) with corresponding AICc scores, Akaike weights (wi), and variation explained

(R2). Null models (i.e., intercept only) are also included. Flood is the magnitude of a 50 AEP flood (a flood that has a 50% chance

of occurring each year). Drainage area is in km2 and channel width is in m. Fire axis is the first principal component describing fire

severity, fire frequency, and fire timing. NDVI is an index of vegetation greenness derived from 1-m2 resolution National

Agriculture Imagery Program (NAIP) remote imagery.

2 Reponse AICc ΔAICc wi R Reliance on aquatically-derived energy (Tetragnathidae)

0 Flood (+), Channel width (-) -96.76 0.00 0.74 0.79

Flood (+), Channel width (-), NDVI (-) -94.62 2.14 0.26 0.79 Null -44.03 52.73 0.00 - Reliance on aquatically-derived energy (benthic macroinvertebrate predators) Flood (+), Channel width (-) -47.28 0.00 0.33 0.23 Drainage area (+), Channel width (-) -46.39 0.89 0.21 0.21 Flood (+) -46.07 1.22 0.18 0.13 Flood (+), Channel width (-), NDVI (+) -44.83 2.45 0.10 0.24 Continued b 220

Table 12 continued Drainage area (+) -43.96 3.32 0.06 0.06 Drainage area (+), Channel width (-), NDVI (+) -43.81 3.47 0.06 0.22 Fire axis (+), Channel width (-) -43.59 3.70 0.05 0.21 Null -41.88 5.40 0.02 - Trophic position (Tetragnathidae) Flood (+), Channel width (-) 16.62 0.00 0.41 0.61 Flood (+) 17.81 1.20 0.23 0.56 Flood (+), NDVI (-) 18.60 1.99 0.15 0.58 Flood (+), Channel width (-), NDVI (-) 18.66 2.05 0.15 0.61

1 Drainage area (+), Channel width (-) 20.41 3.80 0.06 0.56

Null 43.96 27.35 0.00 - Trophic position (benthic macroinvertebrate predators) Channel width (+), NDVI (-) 35.11 0.00 0.49 0.48 Fire axis (-), Channel width (+), NDVI (-) 36.66 1.55 0.23 0.48 Channel width (+), NDVI (-), Drainage area (-) 37.57 2.46 0.14 0.49 Flood (-), Channel width (-), NDVI (+) 37.69 2.59 0.14 0.48 Null 47.53 12.42 0.00 -

b 221

Study reaches

Figure 19. Locations of 31 study reaches along a gradient of drainage area size from the headwaters of the Merced River (grey catchment) and South Fork of the Merced River

(taupe catchment) in the Clark Range to the boundary of Yosemite National Park (YNP),

California, USA. Fire history for the 30 years prior to sampling is also shown (1983-

2012).

222

1 Tetragnathid spiders predatory benthic macroinvertebrates 0.9

0.8

0.7

0.6 derived energy derived - 0.5

0.4

0.3

0.2

Reliance on aquatically on Reliance 0.1

0 0 20 40 60 80 100 120 140 Drainage area (km2)

Figure 20. Reliance on aquatically-derived energy (i.e., proportion of nutritional subsidies derived from benthic algal pathways) by tetragnathid spiders (diamonds) and benthic macroinvertebrate predators (squares) from the upstream end of the Merced and South

Fork of the Merced Rivers to 4th and 5th order mainstem segments near the boundary of

YNP.

223

4.5 Tetragnathid spiders 4 predatory benthic macroinvertebrates

3.5

2.5

1.5 Trophic position Trophic

0.5

0 0 20 40 60 80 100 120 140 Drainage area (km2)

Figure 21. Trophic position of tetragnathid spiders (diamonds) and benthic macroinvertebrate predators (sqaures) from the upstream end of the Merced River and

South Fork of the Merced River to 4th and 5th order mainstem segments near the boundary of YNP.

224

Figure 22. Constructed path diagrams for each response variable: reliance on aquatically- derived energy by (a) tetragnathid spiders (χ2 = 0.01, p = 0.994, CFI = 1.00, TFI = 1.09,

RMSE = 0.00) and (b) predatory benthic macroinvertebrates (χ2 = 5.23, p = 0.514, CFI =

1.00, TFI = 1.01, RMSE = 0.00); trophic position of (c) tetragnathid spiders (χ2 = 0.02, p

225

Flood (a) magnitude R2 = 0.78

0.62 1.07 Reliance on aquatically -0.41 Channel width derived energy (Tetragnathid spiders)

-0.28

(b) 0.17 NDVI 0.12 R2 = 0.26 Channel width

-0.32 Flood Reliance on aquatically 0.78 Magnitude derived energy 1.05 (predatory benthic 1.21 macroinvertebrates) Drainage area -0.55

(c) R2 = 0.60 -0.34 Drainage area 1.21 Flood 1.18 Trophic position 0.78 Magnitude (Tetragnathid spiders) -0.32 Channel width -0.18

-0.19 (d) Drainage area R2 = 0.48 0.92 Fire -0.28 Trophic position 0.78 (predatory benthic macroinvertebrates) -0.43 0.17 NDVI Channel width 0.67

226

5 1200

4 1000

800 )

s 3 3

2 derived energy

- 600

1 Fire axis

400 Trophic positionTrophic 0

0 20 40 60 80 100 120 140 (m magnitude Flood

200 Reliance Reliance onaquatically -1

-2 0 Drainage area (km2) Trophic position Reliance on aquatically-derived energy Fire Flood magnitude

227

0.9 a 0.9 b

0.8 0.8 derived energy derived

- 0.7 0.7

y y

0.6 0.6

0.5 0.5 Reliance on aquatically on Reliance -1 0 1 2 3 0 20 40 60 80 100 4.0 4.0 c x d x

3.5 3.5

3.0 3.0

position

y y

2.5 Trophic 2.5

2.0 2.0

-1 0 1 2 3 0 20 40 60 80 100 Firex PC Drainagex area (km2)

0.001, R2 = 0.75, AIC = -0.15); and (d) Drainage area by trophic position (linear: t =

26.27, p < 0.001, R2 = 0.43, AIC = 25.46; piecewise: t = 35.56, p < 0.001, R2 = 0.70, AIC

= 5.00).

228

Breeanne K. Jackson; S. Mažeika P. Sullivan

Abstract: Wildfire is an important source of disturbance for stream ecosystems of the

American West with demonstrated effects on stream food webs. Thus, wildfire might be expected to influence trophic characteristics of the American dipper (Cinclus mexicanus), a species intimately tied to stream systems for energetic and habitat requirements. We used naturally-abundant stable isotopes of carbon (13C) and nitrogen (15N) to estimate reliance on aquatically-derived energy (i.e., nutritional subsidies derived from aquatic primary production) and trophic position of dippers in mountain streams of the western slope of the central Sierra Nevada in California, USA. This region is influenced by a

Mediterranean-type climate characterized by dry-wet seasonality, high interannual variability in precipitation, and high fire frequency. Model-selection results indicated that larger catchments (greater drainage area) were associated with increased reliance on aquatically-derived energy by dippers indicating the importance of aquatic primary production to terrestrial consumers in larger systems. Precipitation (negative effect), territory length (positive effect), and fire (positive effect) were also supported predictors of reliance on aquatically-derived energy by dippers. Territory length (positive effect)

229 was the most supported predictor of dipper trophic position, although precipitation, fire, and drainage area also received support. Greater proportion of the catchment affected by frequent, recent, or severe wildfire draining to dipper territories was positively related to reliance on aquatically-derived energy by dippers in smaller, headwater streams likely through removal of the conifer canopy and increased light penetration. Conversely, we observed an inverse relationship between precipitation and reliance on aquatically- derived energy by dippers within larger, network streams. We interpret this evidence to indicate that regional climate may be more important than fire for dipper trophic characteristics in larger systems, but that fire may be more important in smaller systems.

Dippers and other aquatic birds spatially and temporally integrate riverine landscapes through their foraging and reproductive activities. Thus, assessment of the trophic dynamics of aquatic-obligate birds may further illuminate both the role of fire as well as other classic environmental determinants of stream-riparin food webs.

230

Introduction

Wildfire is an important source of disturbance in the American West with implications for structural and functional aspects of stream ecosystems (Resh et al. 1988,

Gresswell 1999, Verkaik et al. 2013a). Wildfires can alter stream geomorphology

(Wondzell and King 2003, Shakesby and Doerr 2006), riparian vegetation structure and community composition (Dwire and Kauffman 2003), benthic macroinvertebrate abundance and community composition (Minshall 2003, Verkaik et al. 2013a), and fish populations (Dunham et al. 2003). In addition, the role of wildfire in shaping stream food webs has received increasing attention (Koetsier et al. 2007, Rosenberger et al. 2011,

Harris et al. In revision), where wildfire has been shown to be related to decreased inputs of riparian leaf litter and terrestrial invertebrates to streams (Jackson et al. 2012), increased in-stream primary and secondary productivity and downstream export of invertebrate prey (Koetsier et al. 2007), and increased export of emergent aquatic insects to riparian consumers (Malison and Baxter 2010). Despite the growing interest in the effects of wildfire on stream ecosystems, the influences of wildfire on stream-obligate birds has not been addressed, yet may provide important insights into wildfire-induced shifts in ecosystem function, as river birds have been shown to reflect both the structure and function of fluvial systems (Buckton and Ormerod 2002, Sullivan et al. 2007,

Vaughan et al. 2007, Sullivan and Vierling 2012). Additionally, because riverine birds are often top-consumers in stream food webs, they can reflect processes at lower trophic levels (Steinmetz et al. 2003, Sullivan and Watzin 2008).

231

The American Dipper (Cinclus mexicanus, hereafter, “dipper”) possesses life history traits that make the species a valuable indicator of stream ecosystem condition and function (Ormerod et al. 1991, Logie et al. 1996, Morrissey et al. 2004) and dipper presence has been recommended as a metric suitable for bioassessment of water quality

(Feck and Hall Jr. 2004). Dippers are commonly found year round in mountain streams of the North American West (Ealey 1977, Ormerod 1985). Dippers are highly reliant on aquatic invertebrates as prey but also consume small fish (Ormerod 1985). Dippers utilize structural characteristics of streams (i.e. large wood, overhanging ledges, and boulders) for nesting locations (Kingery 1996). Further, aquatic birds such as dippers may be considered landscape integrators (Sullivan et al. 2007, Vaughan et al. 2007) as they are more mobile than other stream organisms and therefore transport stream-derived nutrients and contaminants both longitudinally and laterally in stream-riparian ecosystems. Finally, adult breeding pairs and their offspring are reliant on a breeding territory that is confined to an established stream segment (Ealey 1977); therefore, during the nesting season dippers reflect environmental condition over a defined spatial extent.

Wildfire has the potential to influence dippers during the reproductive season by affecting structural components of their stream territories, including nesting habitat.

Wildfire can alter the physical characteristic of stream channels if removal of obstructions to overland flow results in increased runoff and erosion (Wondzell and King

2003, Shakesby and Doerr 2006). In some cases, wildfire may result in debris flows that reorganize channels, export large wood, and scour streambeds to bedrock (Miller et al.

2003, Wondzell and King 2003, May 2007). Under these circumstances, nesting locations

232 or activities for dippers may be interrupted or destroyed in the short term (months to years) following fire. However, the creation and destruction of stream habitat that plays out over longer timescales (years to decades) and larger spatial scales (entire catchments) via post-fire fluvial geomorphic processes (Benda et al. 2003) suggest that new nesting locations also may arise over time as a consequence of these same processes.

As a top predator in aquatic systems, dippers may also reflect changes in food- web characteristics following wildfire. For example, benthic macroinvertebrate community composition can shift toward increased abundance of trophic generalists and a decreased abundance of shredders (Mihuc and Minshall 1995). These changes in benthic invertebrate communities are often accompanied by a decrease in overhanging riparian vegetation (Cooper et al. 2014) and an increase in standing crop of epilithic algae

(Minshall et al. 1997, Spencer et al. 2003, Cooper et al. 2014), indicating a shift from food webs based on allochthonous riparian production to autochthonous in-situ production. Therefore, dippers forging in streams affected by wildfire may derive more of their nutritional requirements from aquatic algae (via benthic invertebrate prey) than from stream-conditioned leaf litter (i.e., detritus).

Fire interacts strongly with precipitation patterns to affect aquatic ecosystems

(Gresswell 1999, Arkle et al. 2010, Verkaik et al. 2013b) as post-fire runoff is the primary driver of stream erosion and sedimentation (Shakesby and Doerr 2006).

Observed shifts in stream temperature, benthic algal production, and benthic invertebrate density and community composition following fire primarily depend on channel reorganization (Dunham et al. 2007, Isaak et al. 2010). Whereas extreme precipitation

233 events may be detrimental to dippers by increasing fine sediments and reducing food availability (Price and Bock 1983), moderate increases in precipitation have been linked to increased productivity of dippers (number fledged) (Sullivan and Vierling 2012).

Further, Jackson and Sullivan (In press) observed a consistent signal of precipitation explaining variation in trophic dynamics of riparian spiders independent of fire.

Therefore, precipitation might be expected to play an important role in determining the relative influence of fire on dippers.

Food-chain length (FCL) is an important trophic characteristic, relating to both ecosystem function and stability (Post 2002a). Disturbance – along with ecosystem size and resource availability – is thought to be among the primary drivers of food-chain length (Pimm 2002, Post 2002a, Sabo et al. 2009). For example, Parker and Huryn (2006) found that a spring-fed stream in the Arctic exhibited longer food chains, as exhibited by a presence of dippers, than a similar sized snow-melt fed stream with a flashier hydrograph. Conversely, Thompson and Townsend (1999) found no effect of disturbance

(in this case downstream transport of substrates) on FCL, but did find a positive linear effect of resource availability (organic matter standing crop). In the South Fork Eel River of California, food-web research (Power 1992, 2006, Power et al. 2008) has revealed that in some cases disturbance (i.e., flooding) interacts with individual life histories of organisms, implicating both top-down and bottom-up controls on FCL. Additionally, new evidence implicates hydrologic variability in mediating the effect of ecosystem size on

FCL in rivers (Sabo et al. 2010). Because dippers feed on both insects and small fish, they can be top consumers in mountain stream food webs; thus, dipper trophic position is

234 likely influenced by similar factors as FCL. Most studies of the interaction between disturbance and FCL in streams have focused on flooding and drying (Sabo et al. 2009,

Sabo et al. 2010), however, to our knowledge wildfire has not been examined within this context.

Naturally-abundant stable-isotope analysis has emerged as a valuable technique to describe food webs in aquatic ecosystems because it addresses both trophic position and diet (Collier et al. 2002, Hicks et al. 2005). The trophic position of a consumer organism can be determined from its nitrogen isotope signature as there is a 3-4‰ enrichment of

δ15N (the ratio of 15N to 14N) with each trophic step. In addition, the ratio of 13C to 12C

(δ13C) can vary between terrestrial and aquatic primary producers [i.e., stream algae can exhibit a distinct δ13C from riparian vegetation (e.g., Finlay 2001, Tagwireyi and Sullivan

2015)]. This separation in basal resources is retained in consumer organisms so that the source of a consumer’s energy requirements can be determined from the isotopic signature. Further, dippers that feed on small fish tend to have higher concentrations of contaminants because bioaccumulation and biomagnification of contaminants is higher in fish than it is in invertebrates (Suedel et al. 1994, Kidd et al. 1995). Therefore, wildfire induced shifts in food webs may lead to concomitant shifts in body loading of heavy metals [e.g., mercury (Hg)] in dippers, which together with stable-isotope analysis could further inform our understanding of dipper trophic responses.

Here, we explored the influences of wildfire (i.e., severity, frequency, and timing), ecosystem size, and precipitation on dipper reliance on aquatically-derived energy (i.e., nutritional subsidies derived from benthic algal pathways) and trophic

235 position. We hypothesized that fire would act as a disturbance agent to decrease dipper trophic position via reductions in benthic invertebrate community diversity prompted by shifts in stream hydrogeomorphology. Because fire opens the stream canopy (increasing aquatic primary production) and interacts with precipitation to reorganize stream channels and shift benthic invertebrate community composition toward trophic generalists, we also anticipated that dipper reliance on aquatically-derived energy would increase along a gradient of increasing fire severity, frequency, and timing and that precipitation would interact with fire to increase reliance on aquatically-derived energy.

Given demonstrated differences in the nature of ecological-physical relationships among streams of different sizes (e.g., Sullivan 2012), we expected that relationships between dipper trophic responses and our predictors may be divergent between small headwater streams and larger network systems. In particular, we anticipated that fire might have a more pronounced effect in small headwater streams compared to network streams. For example, narrow riparian zones in headwater streams burn with roughly the same frequency and severity as the upland (Arkle and Pilliod 2010, Van de Water and

North 2010), creating large canopy gaps, whereas in larger systems fire effects in riparian zones are mediated by higher foliar moisture content and relative humidity in comparatively wider, less steep riparian forests (Dwire and Kauffman 2003, Pettit and

Naiman 2007).

236

Methods

Study System

We reconnoitered 27 dipper breeding territories within the Merced and Tuolumne

River basins located within Yosemite National Park, California, USA in 2012 and 2013

(Figure 25). Yosemite National Park of California’s central Sierra Nevada covers 3,027 km2 with 95% of the area designated as wilderness. The region is influenced by the El

Niño Southern Oscillation (ENSO) cycle (DeFlorio et al. 2013). Therefore, although spring snowmelt driven flows are common, precipitation patterns can be highly variable, characterized by fall and winter high-flow events as well as periods of sustained drought.

The regional climate is Mediterranean-type and the park typically receives 94.5 cm of precipitation annually of which 73.7 cm falls as snow (Western Regional Climate Center,

2012). The Tuolumne basin encompasses 5,076 km2 and discharges 70 m3 s-1 on average while the Merced River drains 4,470 km2 and has an average discharge of 34 m3 s-1

(USGS, 2012).

Dippers

We located dipper nests on both the mainstem of the Tuolumne and Merced

Rivers and their tributaries by searching likely rocky overhangs, bridges, boulders, and other prime nesting locations and by following birds while they were constructing nests, incubating, and feeding nestlings (through non-invasive observations). We identified 18 breeding territories in the spring of 2012. In 2013, we identified an additional 9 breeding

237 territories and resampled at 11 breeding territories from the previous season (Table 14,

Figure 25).

The length of each breeding territory, which was used in our analysis as a measure of resource availability, was determined by following dippers both upstream and downstream from the nest until they turned back and recording the waypoint of each territory endpoint using a Garmin eTrex H (Garmin Corporation, Schaffhausen, Canton of Schaffhausen); we observed dippers 2-3 times for each territory endpoint and used the furthest endpoint to calculate territory length following Sullivan and Vierling (2012). We then measured the distance between waypoints along stream corridors using the measure tool in ArcGIS 10.1 (Environmental System Research Institute, Redlands, California,

USA). Although Sullivan and Vierling (2012) used a series of field-based measures using a range finder, we believe our GIS based method would result in little loss of precision and comparable results as precision for our field-based waypoints was typically below

10m, territories in this study were fairly long, and we were able to “trace” territories between waypoints along stream layers in ArcGIS.

At each breeding territory, we assessed habitat quality based on indicators of habitat diversity stemming from the US Environmental Protection Agency’s Rapid

Bioassment Protocols (Barbour et al. 1999). Specifically, we used the Vermont Rapid

Habitat Assessment protocols (RHA; VTDEC 2003), which yields a score from 0 (worst condition) to 200 (reference condition) based on an array of field indicators: woody debris cover, substrate suitable for epifaunal colonization and cover for fish, embeddedness, presence of scour and depositional features, evidence of channel

238 morphology alteration, heterogeneity of velocity and depth combinations, degree to which the channel is filled with water, evidence of bank erosion, cover by riparian vegetation, and occurrence of distinct geomorphic features (i.e., riffles). This method has been modified successfully for use in other mountain areas of the West (i.e., Sullivan and

Vierling 2012). In addition to assessing habitat quality across the reach, higher RHA scores have been widely related to greater benthic macroinvertebrate diversity and abundance (Sullivan et al. 2004, Sullivan and Vierling 2012), therefore we expected

RHA scores to inform our interpretation of dipper trophic dynamics as a proxy for benthic macroinvertebrate community composition and density.

We collected dipper feces from rocks immediately following observed defecation with care taken to avoid collecting any non-fecal material. We also collected feces from below active nests. All samples were collected in the late-spring and early-summer

(between May and July) during incubation and feeding stages and after neslings had fledged, but were still foraging with their parents. Composite fecal samples (one per breeding territory) were frozen until processing in the lab. In addition, as basal resources for our stable isotope analysis, we collected stream-conditioned leaf litter and epilithic/benthic algae from throughout the breeding territory (i.e., upper, middle, and lower; one composite sample per territory). Although blood and/or feathers are commonly used for stable isotope analysis (e.g., Morrissey et al. 2004, Sullivan and

Vierling 2012), the wilderness setting of Yosemite National Park prompted us to consider alternatives. Recently, fecal stable isotopes have been promoted for use in mammalian ecology given their short-term responsiveness to dietary changes (Blumenthal et al. 2012,

239

Salvarina et al. 2013). There has also been some success in using avian feces for stable isotope analysis (δ13C) of captive bird diet (Bird et al. 2008) although Podlesak et al.

(2005) found that feces across a suite of free-living songbirds had the most depleted isotopic values among material/tissue sampled (breath, feathers, plasma, feces). To help resolve potential differences and inform the interpretation of our trophic models using fecal isotope data, we conducted a small study on nesting riverine swallows of the Scioto

River in Columbus, Ohio, USA. Blood and feces samples were collected from 11 individual swallows in the summer of 2014 to assess the relationship between blood and feces relative to stable isotope values of δ13C and δ15N. Detailed methods and results are presented in Appendum B.

Drainage area and precipitation

We delineated the area draining to each breeding territory (i.e., drainage area) using the watershed tool in ArcGIS 10.1 from 10m2 digital elevation model (DEM) geospatial data acquired from the YNP fire atlas and created by the US Geological

Survey and used the downstream-most territory markers as pour points. Drainage area was used as our broad-scale measure of ecosystem size. Following Meyer et al. (2007) and Sullivan (2012), we defined headwater streams as intermittent, first, and second order and network streams as larger (< 3rd order) systems. We gathered precipitation data (cm) from the Western Regional Climate Center, using the closest Remote Automatic Weather

Station (RAWS) or National Weather Service Cooperative Network station (COOP) to each breeding territory that was also situated at approximately the same elevation. Where

240 a nearby station was not available, we used the nearest station within the same catchment

(i.e., Merced River, Tuolumne River, or South Fork of the Merced River). Average monthly precipitation for the water year corresponding to sampling was used to determine precipitation used for analysis [i.e., 01 October of the previous year to 30

September of the sampling year; (Bêche and Resh 2007)]. The total number of precipitation stations utilized was only six, however precipitation was calculated by water-year, so the total number of possible values was twelve. Therefore, not every dipper territory was assigned a unique precipitation value.

Wildfire characteristics

The proportion of each catchment that had been burned in the last decade (fire timing), and greater than two times since 1930 (i.e., fire frequency) was calculated for each breeding territory by using the extract by mask function in ArcGIS Spatial Analyst to pull out fire history data for each delineated catchment (see drainage area methods).

Fire history geospatial data was acquired from the Integrated Resource Management

Applications portal (https://irma.nps.gov/). Burn severity geospatial data was acquired directly from the YNP fire atlas and was estimated for each pixel using normalized burn ratio values (NBR) calculated from Landsat 7 Enhanced Thematic Mapper Satellite

Imagery following (Key and Benson 2006), and relative differences in normalized burn ratios (RdNBR) were calculated from these ratios. We categorized RdNBR values as follows: 0 – unburned, 1 and 2 – low-severity, 3 – moderate-severity, > 4 – high-severity.

We combined moderate and high-severity burned into one category (i.e., moderate-to-

241 high severity burned) and determined the proportion of each catchment falling into this category for every fire > 200 acres occurring after 1984. We then summed the proportion burned with moderate-to-high severity for each fire within each catchment (fire severity).

For fire frequency and severity, we used the full record available from YNP. For fire severity, this record starts in 1984 and for frequency it starts in 1930. We only calculated fire timing from 30 years prior to sampling (1983) because fires occurring within this time period represent over 75% of the total area burned after 1930 and we anticipated that this time period would have the most explanatory power for current ecological signals.

Contaminant and Stable Isotope Analysis

In the laboratory, a portion of each unprocessed dipper fecal sample was sent to the Diagnostic Center for Population and Animal Health at Michigan State University for

Hg analysis. Dipper fecal samples were analyzed for Hg concentrations [μg kg-1 dry weight] at the Diagnostic Center for Population and Animal Health, Toxicology Section,

Michigan State University. We used total Hg as a proxy for its bioaccumulating, organic form methyl mercury (MeHg) because MeHg commonly comprises a high proportion of total Hg in aquatic and riparian consumers (Buckland-Nicks et al. 2014, Speir et al. 2014) and total Hg is often easier to measure as it does not require further speciation. This approach is also consistent with other Hg estimates in dippers (e.g., Morrissey et al.

2004). ICP-AES (Varian Vista, now part of Agilent, Santa Clara, CA) and ICP-MS

(Agilent 7500ce) instruments were calibrated with standards derived from NIST- traceable stock solutions for each element (GFS Chemicals, Inc., Cincinnati, OH).

242

Quality was assured in each sequence run by analyzing lab reagent blanks, NIST

(National Institute of Standards & Technology, Gaithersburg, MD)-traceable Multi-mix

(Alfa Aesar Specpure, Ward Hill, MA) and digests of NIST Standard Reference

Materials (SRM), including Mussel Tissue 2976, Trace Elements in Water 1643e and/or

Montana II Soil 2711a, as appropriate. Mercury was analyzed by cold vapor atomic absorption spectrometry (CETAC CVAA, Omaha, NB) with similar calibration and quality control.

The remaining dipper feces samples were dried in a 60o C oven for at least 48 hours and subsequently homogenized using a mortar and pestle. Benthic algae and stream-conditioned leaf litter were sorted from sediments and invertebrates. Samples were then rinsed with distilled water, dried at 60o C for 48 hours, and then homogenized into a fine powder using a Pica Blender Mill (Cianflone Scientific Instruments

Corporation, Pittsburgh, Pennsylvania, USA) or mortar and pestle. All samples (feces, detritus, and algae) were weighed and packed into tin capsules for stable isotope analysis.

Continuous flow isotope-ratio mass spectrometry (EA-IRMS) was used to analyze all samples for 13C and 15N by at the Washington State University Stable Isotope Core

(Pullman, Washington, USA). The results are reported in δ (‰) notation defined as:

13 15 δ C or δ N = [(Rsample/Rstandard)-1] * 100 where R is 13C/12C or 15N/14N, respectively. Typical analytical precision was 0.08‰ for

δ15N and 0.19‰ for δ 13C determination.

243

Trophic position and reliance on aquatically-derived energy

We applied a correction factor to δ13C and δ15N values of dipper feces based on linear regression of paired blood and feces samples from the Ohio swallows study (see

Appendum B). To estimate dipper trophic position (TP), we used the two-source food web model from Post (2002b); TP = λ + { δc – [ δb1 * α + δb2 * (1-α)]}/ Δn where λ is the trophic position of the basal food sources (i.e., 1 for primary producers); δc is

15 the δ N signature of the consumer; δb1 and δb2 are the signatures of the two basal food sources; α is the proportion of N from basal food source 1; and Δn is the enrichment in

δ15N per trophic level (i.e., 3.4‰; Post 2002b). A two-end member Bayesian isotopic mixing model was used to determine the proportion of N derived from basal source 1

(i.e., α) with the R software package SIAR (Stable Isotope Analysis in R; Parnell and

Jackson 2013). Both δ13C and δ15N, data were used to estimate the contribution from each basal food source to dippers. We estimated the contribution from each basal food source (i.e., detritus or algae) to dippers using a two-end member Bayesian isotopic mixing model also using SIAR in R.

Statistical analysis

We used paired t-tests to test for interannual variability in reliance on aquatically- derived energy and trophic position of dippers. We performed principle component analysis (PCA) on fire frequency, severity, and timing variables extracted from our geospatial data. The first PC represented 69.1% of the variance in the fire dataset (Table

13) and described a gradient of frequent, recent, and severe fire within each catchment.

244

We examined the relative explanatory power of the fire axis (as a measure of disturbance), drainage area (as a measure of ecosystem size), territory length (as a measure of resource availability), and precipitation (as a measure of climate) on dipper trophic position and reliance on aquatically-derived energy. To do this we used an information-theoretic model selection approach based on Akaike’s information criterion

(AIC) (Anderson and Burnham 2002, Burnham and Anderson 2004). Competing models were selected per response variable by linear regression. Akaike Information Criterion adjusted for small sample size (AICc) was calculated for each competing model and models with ΔAICc ≤ 4 were retained as the most highly supported models (Burnham and Anderson 2002). We determined the relative evidence that a model was the best supported among all of the candidate models in the set by comparing Akaike weights

(ωi). The null model (i.e., intercept only) was also included in each of the sets of competing models for comparative purposes. Only independent variables that were not highly correlated (i.e., r < 0.80) were used together in model selection. Highly-correlated independent variables were assessed in separate models to avoid problems with colinearity. We performed a simple linear regression to assess the link between trophic position and Hg burden. We also used linear regression to test for potential relationships between dipper reliance on aquatically-derived energy and trophic position in network and headwater streams. For all regression analyses we used α = 0.05 to indicate statistical significance and α = 0.10 as a trend (e.g., Rowse et al. 2014). All procedures were performed with JMP 11 software (SAS Institute Inc., Cary, North Carolina, USA).

245

Results

Dipper breeding territories were located on 1st- through 5th-order streams representing mainsteam and/or tributaries of river systems in YNP, and ranged in elevation from 1,074 to 2,373 m (Table 14, Figure 25). Drainage area ranged from 0.93 km2 to 83.13 km2 (Table 15). All headwater streams were < 8 km2 and all network streams > 9 km2. The highest proportion of each catchment influenced by fire was burned once, with low-severity, in the last 30 years ( = 16.4% ± 15.6% (SD), 19.0% ± 22.4%

(SD), and 21.1% ± 24.0% (SD) respectively; Table 15). Dipper territory length was highly variable and ranged from 246 to 1,664 m. Rapid Habitat Assessment scores were fairly consistent across dipper breeding territories and corresponding condition was either good or reference (Table 14). Rapid Habitat Assessment scores were negatively related to dipper territory length (R2 = 0.17, F = 5.41, p = 0.045) (Figure 26).

Across all dipper territories, mean δ13C was -22.21‰ ± 5.27‰ (SD) for epilithic algae, -27.25‰ ± 1.06‰ (SD) for stream-conditioned leaf litter, and -24.60‰ ± 1.94‰

(SD) for dipper feces. Mean δ15N was -1.93‰ ± 1.81 ‰ (SD) for epilithic algae, -2.06‰

± 1.77 ‰ (SD) for stream-conditioned leaf litter, and 0.70‰ ± 1.45‰ (SD) for dipper feces. After applying a correction factor to δ13C and δ15N values for dipper feces, mean

δ13C was -24.51‰ ± 0.96‰ (SD) and mean δ15N was 9.79‰ ± 0.55‰ (SD). Differences in stable-isotope ratios of C and N were sufficient between basal sources to run two- source mixing models (Post 2002b).

Reliance on aquatically-derived energy by dippers ranged from 0.33 to 0.93 with more values falling on the lower end of the range (i.e., median = 0.58). Corrected

246 estimates of trophic position spanned almost one trophic level (i.e., 4.19 to 4.96) (Table

15). Paired t-tests of reliance on aquatic-energy sources and trophic position of dippers between 2012 and 2013 sampling years revealed no significant difference between years

(Δ = 0.03, t = 1.30, df = 10, p = 0.223; Δ = 0.12, t = 1.58, df = 10, p = 0.144, respectively). Trophic position was not significantly related to Hg concentration in feces

(p > 0.05, data not shown). Mean Hg concentration was 94.44 μ kg-1 ± 40.86 μ kg-1 (SD) and ranged from 40 to 173 μ kg-1.

None of the predictor variables used in model selection were highly correlated

(i.e., r = -0.09 to 0.46). Model selection showed that drainage area, fire, and precipitation were important predictors of reliance on aquatically-derived energy by dippers and that territory length also played a role (Table 16). Drainage area (ωi = 0.43) alone explained

48% of the variation in reliance on aquatically-derived energy, and together drainage area and precipitation explained 52% of the variation observed (ωi = 0.51). Models including fire and drainage area and fire, drainage area, and precipitation also received support (ωi

= 0.49 and ωi = 0.53, respectively). All predictor variables resulted in supported models for dipper trophic position, but a univariate model including territory length (ωi = 0.30) was the only model with a lower AICc score than the null model (Table 16). Territory length alone explained 13% of the variation. A model containing trophic position with drainage area, precipitation, or fire also received weak support (ωi = 0.14, ωi = 0.11, and

ωi = 0.07; respectively). Univariate models including drainage area, fire, or precipitation were within 4 AICc units from the best-supported model, but received minimal support as predictors of dipper trophic position (ωi = 0.06 for all) (Table 16).

247

Precipitation was inversely related to reliance on aquatically-derived energy by dippers in network streams (i.e., drainage area > 9 km2) (Figure 27a, p = 0.049), but not headwater streams (Figure 27a, p > 0.05). Fire showed a strong positive relationship with reliance on aquatically-derived energy by dippers in headwater streams (Figure 27e, p =

0.001), but not in network streams (Figure 27e). Dipper trophic position was higher for dippers occupying longer territories in network streams (Figure 27d, p = 0.101) but not headwater streams (Figure 27d). Relationships between the fire axis and trophic position

(p > 0.05) and reliance on aquatically-derived energy (p > 0.05) were not significant for either size stream system (Figure 27e,f).

Discussion

In this study, we examined the relationships between trophic characteristics of the

American dipper and environmental variability including precipitation, wildfire, resource availability, and ecosystem size. Ecosystem size, as measured by drainage area received support as a predictor of both reliance on aquatically-derived energy and trophic position.

Precipitation and fire also received model support as complementary predictors of dipper nutritional pathways, but not for trophic position. Whereas territory length was the best supported predictor of trophic position (positive relationship).We also present initial evidence that the nature of trophic responses of dippers may be different in headwater versus network streams. Few studies have considered organisms like aquatic birds that spatially integrate riverine landscapes as indicators of stream-riparian food webs (Fausch et al. 1997, Sullivan et al. 2007, Sullivan and Vierling 2012), and none have examined

248 wildfire as a potential driver of dipper trophic responses. Thus, this study is an important first step in describing wildfire effects on an aquatic-obligate bird.

Reliance on aquatically-derived energy

We expected dippers occupying breeding territories located in catchments characterized by more frequent, severe, or recent wildfire to exhibit greater reliance on aquatically-derived energy. Although we found support for fire as an explanatory variable, the weight of evidence was stronger for drainage area and precipitation, which collectively explained more of the variation observed in reliance on aquatically-derived energy (Table 16). However in headwater streams, the fire axis had a strong positive relationship with reliance on aquatically-derived energy by dippers (Figure 27e) suggesting that a greater proportion of the catchment affected by recent, frequent, or severe fire results in a greater proportion of the food chain leading to dippers to be supported by aquatically-derived energy (i.e., benthic algae), likely driven by opening of the riparian canopy by fire (Dwire and Kauffman 2003), increased in-stream primary productivity (Betts and Jones 2009), and increased abundance of grazers (Mihuc and

Minshall 2005). We also found that dippers relied more heavily on aquatically-derived energy in larger catchments, which aligns with theoretical and empirical evidence that headwater food webs are more heavily influenced by allochthony (i.e., inputs of organic matter from the riparian zone) than mid-order stream food webs that shift toward greater autochthony (i.e., primary production from in stream autotrophs) (Vannote et al. 1980,

Wallace et al. 1997, Rosi-Marshall and Wallace 2002, Hu et al. 2005, Protasov 2008).

249

Taken together these results provide initial evidence that although, in general, allochthony is more dominant in smaller streams, fire can shift these systems toward greater authochthony and increase the provisioning of riparian consumers by aquatic primary production.

Precipitation emerged as a salient predictor of reliance on aquatically-derived energy by dippers in two supported models (Table 16). Dippers require clear unpolluted water for foraging and typically rely on benthic macroinvertebrates classically found to be intolerant of poor chemical water quality such as Ephemeroptera, Trichoptera, and

Plecoptera (Ormerod et al. 1991). In ecosystems affected by Mediterranean-type climate, as in our central Sierra Nevada study location (Bonada and Resh 2013), high interannual variability in precipitation may have a greater influence on food resources available to dippers (i.e., benthic invertebrate prey) than fire (Verkaik et al. 2013b). For example,

Bêche and Resh (2007) found that benthic invertebrate communities in the north-central

Sierra Nevada shifted from drought-adapted communities in dry years to communities adapted to high flows in wet years. These shifts were predominantly due to increased dominance of Chironomidae that led to decreased diversity in dry years despite increased richness. In addition, these investigators found that small stream invertebrate communities were more stable under drought conditions than larger stream communities.

The influence of precipitation on reliance on aquatically-derived energy by dippers in network streams collaborates this result, and points to dry-year shifts in benthic invertebrate prey (especially in network streams) as a potential driver of trophic shifts in dippers. However, exactly how shifts in the composition of available benthic prey could

250 be influencing reliance on aquatically-derived energy by dippers is not apparent from our study. Although increased precipitation might seem to interact with fire leading to increased overland flow and sedimentation (Shakesby and Doerr 2006), thereby resulting in reduced benthic macroinvertebrate diversity and increased dominance by trophic generalists (Robinson et al. 2000, Minshall et al. 2001, Mihuc and Minshall 2005, Arkle et al. 2010), in systems influenced by Mediterranean-type climate, drought and low-flow scenarios may be more important for determining benthic macroinvertebrate community composition (Bêche and Resh 2007, Verkaik et al. 2013b) (and therefore dipper trophic dynamics) than has been observed in temperate systems (Verkaik et al. 2013a).

Trophic position

Using stable-isotope analysis of dipper feces, trophic position estimates ranged from 4.19 to 4.96 in our Yosemite study, which is higher than dipper trophic position reported from northern Idaho streams (3.11, ranging from 2.47 to 3.99; Sullivan et al. In revision). However, our study system included larger streams (up to 5th order compared to a 3rd-order maximum in Sullivan et al. In revision), which are known to support longer food chains (see following paragraph). Previous studies have used blood or feathers for stable-isotope analysis (Morrissey et al. 2004, Sullivan and Vierling 2012), and δ15N values from these studies are between 5-13‰: δ15N was substantially more depleted in our study with only 6 samples >2‰. Isotopic composition of riverine swallow feces was also significantly more depleted than blood (Appendum B). Following application of a correction factor (i.e., extrapolation of the linear regression between swallow blood and

251 feces samples onto dipper feces samples), our estimates for δ15N were closer to those reported in the literature (i.e., 9-12‰) although still at the high end of the range, and therefore should be interpreted with caution when compared with values from other studies.

Territory length received the greatest weight of evidence as a predictor of dipper trophic position, whereby dippers on longer territories exhibited elevated trophic position.

The longest territories in our study were much longer than those reported by others (i.e.,

Sullivan and Vierling 2012), however, in our study system some dipper territories were located along stream reaches with long sections of exposed bedrock. Therefore dippers may have retained longer territories to compensate for these long sections of stream likely characterized by minimal foraging potential. The observed inverse relationship between territory length and RHA score, indicating that dippers retained longer territories where habitat quality was diminished (even slightly), collaborates this explanation.

Similarly, dippers on longer territories may have been buffered from low-flow shifts in composition of benthic invertebrate prey (see reliance on aquatically-derived energy section above) allowing them to occupy a higher trophic position even during low- precipitation years.

Food-web theory predicts that FCL (and trophic position of top consumers) will be positively related to ecosystem size (Pimm 2002, Post 2002a, Sabo et al. 2009), wherein larger ecosystems will have longer food chains via a suite of mechanisms

(reviewed in Warfe et al. 2013), including because they support more basal resources and have greater species richness (Schoener 1989, Cohen and Newman 1991). However,

252 empirical evidence has yielded mixed results in stream ecosystems (McHugh et al. 2010,

Sabo et al. 2010, Warfe et al. 2013). Morrisey et al (2004) found that dippers occupying breeding territories on large, mainstem stream systems were more enriched in 15N than those occupying breeding territories on smaller tributaries, and this was primarily due to an increased proportion of fish in their diet. A similar mechanism may be at work in our system where dipper trophic position increased with drainage area (and thus stream size, although note that models supporting drainage area received less support than the null model, Table 16). Sullivan and Vierling (2012) found that breeding season precipitation exerted positive influences on dipper δ15N, implicating higher dipper trophic position, and the addition of precipitation to a territory length model added explanatory power (but note that this model was also less well supported than the null model; Table 16).

Wildfire effects

The importance of fire for determining dipper reliance of aquatically-derived energy and trophic position received minimal support in model selection, however the strong positive relationship between fire and reliance on aquatically-derived energy in headwater streams suggests that these relationships may differ by ecosystem size or position in the drainage network. We would expect fire to have a greater influence in smaller streams leading to greater reliance on aquatically derived energy: In smaller stream systems, riparian canopy opened by fire would have a disproportionate effect on light penetration to streams potentially leading to increased primary productivity by benthic algae (i.e., Mihuc and Minshall 2005) and reduced inputs of terrestrial leaf litter

253

(Jackson et al. 2012) leading to a shift from allochthony to autochthony in stream food webs.

The vast majority of research relating fire to stream ecosystem responses has been conducted in headwater streams (Gresswell 1999) and few have utilized highly mobile organisms such as birds that may better integrate catchment-level environmental characteristics (Saab 1999, Steinmetz et al. 2003, Vaughan et al. 2007, Sullivan and

Vierling 2012). Although the effects of fire on stream food webs may be short-lived at smaller spatial scales (Gresswell 1999), over longer timescales fire may continue to play a role in reorganizing downstream habitat (May and Gresswell 2003, Miller et al. 2003,

Wondzell and King 2003, Mayor et al. 2007), and observations of dippers may better illuminate these patterns. Sampling among catchments that have a greater proportion of recent, frequent, or severe fire might yield more pronounced responses (e.g., Arkle and

Pilliod 2010). In our study the mean proportion of each catchment burned in the five years prior to sampling, with greater frequency than 2, and with moderate-to-high severity were all fairly minimal (Table 15).

As terrestrial organisms, dipper trophic position may also not respond to the same suite of environmental determinants as aquatic consumers (i.e., fish, which are typically measured in aquatic FCL studies, for example). Dippers may be released from the effects of some of the classic mechanisms related to FCL like productivity because unlike other top consumers in aquatic ecosystems, they are not restricted to aquatic habitat and can integrate environmental conditions over broader spatial scales. Warfe et al. (2013) suggest that aquatic waterbirds may occupy a higher trophic level than piscivorous fish,

254 and are less likely to respond to local environmental determinants. Dippers during the reproductive season are constrained by territory length, but we found that longer territories were positively (although weakly) related to trophic position, indicating that behavioral mechanisms [i.e., interspecific competition for breeding territories, foraging strategies (Hashiguchi and Yamagishi 1981, Eguchi 1988, Marzolin 2002, Willson and

Hocker 2008)] may be an important factor influencing dipper trophic position. Territory lengths in our study were also negatively related to habitat quality (and potentially aquatic invertebrate food sources; Sullivan and Watzin 2008, Sullivan 2012) (Figure 2), suggesting that, similar to other studies (e.g., Chen and Wang 2010) dippers require shorter territories where in-stream habitat is better, but that their trophic position will be lower (Table 16).

Hg levels were slightly lower in our study compared to those reported for other systems (e.g., 40 – 173 μg kg-1 compared to 90 – 190 μg kg-1 in British Columbia,

Canada; Morrissey et al. 2004) and below levels known to cause toxicity (Beyer et al.

1996). Other studies have implicated increased trophic position, specifically increased proportion of dipper diet comprised of fish (Morrissey et al. 2004) with associated increases in total Hg burden, however we did not find a significant relationship between

Hg concentration in feces and trophic position. Our sample size was relatively low (n =

18), so additional replicates might strengthen our ability to detect a relationship between trophic position and Hg concentration.

255

Conclusions

Dippers are highly mobile organisms that integrate environmental factors over a range of spatial scales (Sullivan and Vierling 2012), therefore we might expect dipper trophic dynamics to be less constrained than aquatic organisms by disturbances like fire and influenced to a greater degree by regional climate as hinted at by the relationships between precipitation and dipper trophic responses observed in our Yosemite study. The distinctive dynamics of wet-dry systems has not been investigated broadly relative to trophic dynamics in fluvial systems. In the wet-dry tropics of northern Australia, Warfe et al. (2013) found no significant link between productivity, habitat quality, or disturbance and FCL and speculate that annual inundations, although short-lived, allow a high-degree of fish dispersal that stabilizes FCL across gradients of productivity, habitat quality, and disturbance with trophic plasticity implicated as a major driver (Polis et al. 1997,

McCann and Rooney 2009, Jardine et al. 2012). Although we found that ecosystem size was a strong predictor of dipper trophic responses, variability in trophic position and reliance on aquatically-derived energy appeared to be somewhat invariant relative to fire

(except for in small, headwater streams), suggesting that dippers are resilient, perhaps through foraging plasticity or habitat selection. Most research into the effects of wildfire on stream ecosystems has focused on relatively less-mobile organisms that experience their environment at smaller spatial scales; because dippers are highly mobile organisms that integrate environmental condition over space and time, further inquiry into the thresholds at which dippers respond to wildfire may have unique implications for

256 managing wildfire beyond the reach scale (e.g., Frissell et al. 1986, Poff 1997, Saab

1999).

Acknowledgements

Funding was provided by NSF DEB-1401480 awarded to SMPS and BKJ, Bureau of

Land Management (14-3-01-37) award to SMPS, and The Ohio State University. In

Yosemite National Park we received help from Dr. G. Smith, S. Stock, J. Roche, and K.

Van Wagtendonk. We appreciate field and laboratory assistance received from D. Groff,

M. Hickson, Dr. K. Hossler, P. Koubek, M. Ledford, and L. Meyer.

257

Literature Cited

Anderson, D. R. and K. P. Burnham. 2002. Avoiding pitfalls when using information- theoretic methods. Journal of Wildlife Management 66:912-918. Arkle, R. S. and D. S. Pilliod. 2010. Prescribed fires as ecological surrogates for wildfires: A stream and riparian perspective. Forest Ecology and Management 259:893-903. Arkle, R. S., D. S. Pilliod, and K. Strickler. 2010. Fire, flow and dynamic equilibrium in stream macroinvertebrate communities. Freshwater Biology 55:299-314. Bêche, L. A. and V. H. Resh. 2007. Short-term climatic trends affect the temporal variability of macroinvertebrates in California 'Mediterranean' streams. Freshwater Biology 52:2317-2339. Benda, L., D. Miller, P. Bigelow, and K. Andras. 2003. Effects of post-wildfire erosion on channel environments, Boise River, Idaho. Forest Ecology and Management 178:105-119. Beyer, W. N., G. H. Heinz, and A. W. Redmon-Norwood. 1996. Environmental contaminants in wildlife: interpreting tissue concentrations. CRC Press, Boca Raton, FL. Bird, M. I., E. Tait, C. M. Wurster, and R. W. Furness. 2008. Stable carbon and nitrogen isotope analysis of avian uric acid. Rapid Communications in Mass Spectrometry 22:3393-3400. Betts, E.F. and J.B. Jones, Jr. 2009. Impact of wildfire on stream nutrient chemistry and ecosystem metabolism in boreal forest catchments of interior Alaska. Arctic and Alpine Research. 41:407-417. Blumenthal, S. A., K. L. Chritz, J. M. Rothman, and T. E. Cerling. 2012. Detecting intraannual dietary variability in wild mountain gorillas by stable isotope analysis of feces. Proceedings of the National Academy of Sciences of the United States of America 109:21277-21282. Bonada, N. and Resh, V.H. 2013. Mediterranena-climate streams and rivers: geographically separated but ecologically comparable freshwater systems. Hydrobiologia 719: 1-29. Buckland-Nicks, A., K. N. Hiller, and N. J. O'Driscoll. 2014. Mercury bioaccumulation in dragonflies (Odonata: Anisoptera): examination of life stages and body regions. Environmental Toxicology and Chemistry 2014. Buckton, S. T. and S. J. Ormerod. 2002. Global patterns of diversity among the specialist birds of riverine landscapes. Freshwater Biology 47:695-709. Burnham, K. P. and D. R. Anderson. 2004. Multimodel inference - understanding AIC and BIC in model selection. Sociological Methods & Research 33:261-304.

258

Chen, C.-C. and Y. Wang. 2010. Relationships between stream habitat and breeding territory length of the Brown Dipper (Cinclus pallasii) in Taiwan. Journal of Ornithology 151:87-93. Cohen, J. E. and C. M. Newman. 1991. Community area and food chain length: theoretical predictions. American Naturalist 138:1542-1554. Collier, K. J., J. S. Bury, and M. Hibbs. 2002. A stable isotope study of linkages between stream and terrestrial food webs through spider predation. Freshwater Biology 47:1651-1659. Cooper, S. D., H. M. Page, S. W. Wiseman, K. Klose, D. Bennet, T. Even, S. Sadro, C. E. Nelson, and T. L. Dudley. 2014. Physicochemical and biological responses of streams to wildfire severity in riparian zones. Freshwater Biology. DeFlorio, M. J., D. W. Pierce, D. R. Cayan, and A. J. Miller. 2013. Western US extreme precipitation events and their relation to ENSO and PDO in CCSM4. Journal of Climate 26:4231-4243. Dunham, J. B., A. E. Rosenberger, C. H. Luce, and B. E. Rieman. 2007. Influences of wildfire and channel reorganization on spatial and temporal variation in stream temperature and distribution of fish and amphibians. Ecosystems 10:335-346. Dunham, J. B., M. K. Young, R. E. Gresswell, and B. E. Rieman. 2003. Effects of fire on fish populations: landscape perspectives on peristance of native fishes and nonnative fish invasion. Forest Ecology and Management 178:183-196. Dwire, K. A. and J. B. Kauffman. 2003. Fire and riparian ecosystems in landscapes of the western USA. Forest Ecology and Management 178:61-74. Ealey, D. M. 1977. Aspects of the ecology and behavior of a breeding population of dippers (Cinclus mexicanus: Passeriformes) in southern Alberta., Edmonton, Alberta, CA. Eguchi, K. 1988. The ecological significant of time budgets of the brown dipper Cinclus- pallasii temminck. Japanese Journal of Ecology 38:243-256. Fausch, K. D., S. Nakano, and S. Kitano. 1997. Experimentally induced foraging mode shift by sympatric charrs in a Japanese mountain stream. Behavioral Ecology 8:414-420. Feck, J. and R. O. Hall Jr. 2004. Response of American dippers (Cinclus mexicanus) to variation in stream water quality. Freshwater Biology 49:1123-1137. Finlay, J. C. 2001. Stable-carbon-isotope ratios of river biota, implications for energy flow in lotic food webs. Ecology 82:1052-1064. Frissell, C. A., W. J. Liss, C. E. Warren, and M. D. Hurley. 1986. A hierarchical framework for stream habitat classification: viewing streams in a watershed context. Environmental Management 10:199-215.

259

Gresswell, R. E. 1999. Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193-221. Harris, H. E., C. V. Baxter, and J. M. Davis. In revision. Wildfire and debris flows affect prey subsidies with implications for riparian and reverine predators. Freshwater Science. Hashiguchi, D. and S. Yamagishi. 1981. Dispersion pattern and territorial behavior of the wintering brown dipper Cinclus-pallasiiI. Japanese Journal of Ecology 31:161- 170. Hicks, J. H., M. S. Wipfli, D. W. Lang, and M. E. Lang. 2005. Marine-derived nitrogen and carbon in freshwater-riparian food webs of the Copper River Delta, southcentral Alaska. Oecologia 144:558-569. Hu, B., L. Yang, B. Wang, and L. Shan. 2005. Functional feeding groups of macroinvertebrates in 1 - 6 order tributaries of the Changjiang River. Chinese Journal of Applied and Environmental Biology 11:463-466. Isaak, D. J., C. H. Luce, B. E. Rieman, D. E. Nagel, E. E. Peterson, D. L. Horan, S. Parkes, and G. L. Chandler. 2010. Effects of climate change and wildfire on stream temperatures and salmonid thermal habitat in a mountain river network. Ecological Applications 20:1350-1371. Jackson, B. K., S. M. P. Sullivan, and R. L. Malison. 2012. Wildfire severity mediates fluxes of plant material and terrestrial invertebrates to mountain streams. Forest Ecology and Management 278:27-34. Jardine, T. D., B. J. Pusey, S. K. Hamilton, N. E. Pettit, and P. M. Davies. 2012. Fish mediate high food web connectvity in the lower reaches of a tropical floodplain river. Oecologia 168:829-838. Key, C.H. and N.C. Benson. 2006. Landscape assessment: ground measure of severity, the Composite Burn Index; and remote sensing of severity, the Normalized Burn Ratio. In D.C. Lutes; R.E. Keane; J.F. Caratti; C.H. Key; N.C. Benson; S. Sutherland; and L.J. Gangi. 2006. FIREMON: Fire effects monitoring and inventory system. USDA Forest Service, Rocky Mountain Research Station, Ogden, UT. Gen. Tech. Rep. RMRS-GTR-164-CD: LA 1-51. Kidd, K. A., D. W. Schindler, R. H. Hesslein, and D. C. G. Muir. 1995. Correlation between stable nitrogen isotope ratios and concentrations of organochlorides in biota from a fresh-water food-web. Science of the Total Environment 160-61:381- 390. Kingery, H. E. 1996. American Dipper (Cinclus mexicanus).in A. Poole and F. Gill, editors. The Birds of North America. The Academy of Natural Sciences and The American Ornithologists Union, Philadelphia, P.A. and Washington, D.C. Koetsier, P., Q. Tuckett, and J. White. 2007. Present effects of past wildfires on the diets of stream fish. Western North American Naturalist 67:429-438.

260

Logie, J. W., D. M. Bryant, D. L. Howell, and J. A. Vickery. 1996. Biological significance of UK critical load exceedance estimates for flowing waters: Assessments of dipper Cinclus cinclus populations in Scotland. Journal of Applied Ecology 33:1065-1076. Malison, R. L. and C. V. Baxter. 2010. The fire pulse: wildfire stimulates flux of aquatic prey to terrestrial habitats driving increases in riparian consumers. Canadian Journal of Fisheries and Aquatic Sciences 67:570-579. Marzolin, G. 2002. Influence of the mating system of the Eurasian dipper on sex-specific local survival rates. Journal of Wildlife Management 66:1023-1030. May, C. 2007. Sediment and wood routing in steep headwater streams: An overview of geomorphic processes and their topographic signatures. Forest Science 53:119- 130. May, C. L. and R. E. Gresswell. 2003. Processes and rates of sediment and wood accumulation in headwater streams of the Oregon Coast Range, USA. Earth Surface Processes and Landforms 28:409-424. Mayor, A. G., S. Bautista, J. Llovet, and J. Bellot. 2007. Post-fire hydrological and erosional responses of a Mediterranean landscape: seven years of catchment-scale dynamics. Catena 71:68-75. McCann, K. S. and N. Rooney. 2009. The more food webs change, the more they stay the same. Philos T R Soc B 364. McHugh, P. A., A. R. McIntosh, and P. G. Jellyman. 2010. Dual influences of ecosystem size and disturbance on food chain length in streams. Ecology Letters 13:881-890. Meyer, J. L., D. L. Strayer, J. B. Wallace, S. L. Eggert, G. S. Helfman, and N. E. Leonard. 2007. The contribution of headwater streams to biodiversity in river networks. Journal of the American Water Resources Association 43:86-103. Mihuc, T. B. and G. W. Minshall. 1995. Trophic generalists vs trophic specialists - implications for food-web dynamics in post-fire streams. Ecology 76:2361-2372. Mihuc, T. B. and G. W. Minshall. 2005. The trophic basis of reference and post-fire stream food webs 10 years after wildfire in Yellowstone National Park. Aquatic Sciences 67:541-548. Miller, D., C. Luce, and L. Benda. 2003. Time, space, and episodicity of physical disturbance in streams. Forest Ecology and Management 178:121-140. Minshall, G. W. 2003. Responses of stream benthic macroinvertebrates to fire. Forest Ecology and Management 178:155-161. Minshall, G. W., C. T. Robinson, and D. E. Lawrence. 1997. Postfire responses of lotic ecosystems in Yellowstone National Park, USA. Canadian Journal of Fisheries and Aquatic Sciences 54:2509-2525.

261

Minshall, G. W., C. T. Robinson, D. E. Lawrence, D. A. Andrews, and J. T. Brock. 2001. Benthic macroinvertebrate assemblages in five central Idaho (USA) streams over a 10-year period following disturbance by wildfire. International Journal of Wildland Fire 10:201-213. Morrissey, C. A., L. I. Bendell-Young, and J. E. Elliot. 2004. Linking contaminant profiles to the diet and breeding locations of American dippers using stable isotopes. Journal of Applied Ecology 41:502-512. Ormerod, S. J. 1985. The diet of breeding dippers (Cinclus, mexicanus) and their nestlings in the catchment of the River Wye, mid-Wales: a preliminary study by faecal analysis. Ibis 127:316-331. Ormerod, S. J., J. O. Halloran, S. D. Gribbin, and S. J. Tyler. 1991. The ecology of dippers Cinclus cinclus in relation to stream acidity in upland wales: breeding performance, calcium physiology and nestling growth. Journal of Applied Ecology 28:419-433. Parker, S. M. and A. D. Huryn. 2006. Food web structure and function in two arctic streams with contrasting disturbance regimes. Freshwater Biology 51:1249-1263. Parnell, A. and A. Jackson. 2013. Stable isotope analysis in R. http://cran.r- project.org/web/pakage/siar/. Pettit, N. E. and R. J. Naiman. 2007. Fire in the riparian zone: characteristics and ecological consequences. Ecosystems 10:673-687. Pimm, S. L. 2002. Food webs. The University of Chicago Press, Chicago. Podlesak, D. W., S. R. McWilliams, and K. A. Hatch. 2005. Stable isotopes in breath, blood, feces and feathers can indicate intra-individual changes in the diet of migratory songbirds. Oecologia 142:501-510. Poff, L. N. 1997. Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16:391-409. Polis, G. A., W. B. Anderson, and R. D. Holt. 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Review of Ecology and Systematics 28:289-316. Post, D. M. 2002a. The long and short of food-chain length. Trends in Ecology & Evolution 17:269-278. Post, D. M. 2002b. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83:703-718. Power, M. E. 1992. Habitat heterogeneity and the functional significance of fish in river food webs. Ecology 73:1675-1688. Power, M. E. 2006. Environmental controls on food web regimes: a fluvial perspective. Progress in Oceanography 68:125-133.

262

Power, M. E., M. S. Parker, and W. E. Dietrich. 2008. Seasonal reassembly of a river food web: floods, droughts, and impacts of fish. Ecological Monographs 78:263- 282. Price, F. E. and C. E. Bock. 1983. Population ecology of the dipper (Cinclus mexicanus) in the Front Range of Colorado. Studies in Avian Biology:1-84. Protasov, A. A. 2008. River and lake continua: an attempt at analysis and synthesis. Inland Water Biology 1:105-113. Resh, V. H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W. Li, G. W. Minshall, S. R. Reice, A. L. Sheldon, J. B. Wallace, and R. C. Wissmar. 1988. The role of disturbance in stream ecology. Journal of the North American Benthological Society 7:433-455. Robinson, C. T., G. W. Minshall, and T. V. Royer. 2000. Inter-annual patterns in macroinvertebrate communities of wilderness streams in Idaho, USA. Hydrobiologia 421:187-198. Rosenberger, A. E., J. B. Dunham, J. M. Buffington, and M. S. Wipfli. 2011. Persistent effects of wildfire and debris flows on the invertebrate prey base of Rainbow Trout in Idaho streams. Northwest Science 85:55-63. Rosi-Marshall, E. J. and J. B. Wallace. 2002. Invertebrate food webs along a stream resources gradient. Freshwater Biology 47:129-141. Rowse, L. M., A. D. Rodewald, and S. M. P. Sullivan. 2014. Pathways and consequences of contaminant flux to Acadian flycatchers (Empidonax virescens) in urbanizing landscapes of Ohio, USA. Science of the Total Environment 485:461-467. Saab, V. 1999. Importance of spatial scale to habitat use by breeding birds in riparian forests: a hierarchical analysis. Ecological Applications 9:135-151. Sabo, J. L., J. C. Finlay, T. Kennedy, and D. M. Post. 2010. The role of discharge variation in scaling of drainage area and food chain length in rivers. Science 330:965-967. Sabo, J. L., J. C. Finlay, and D. M. Post. 2009. Food chains in freshwaters. Pages 187- 220 in R. S. Ostfeld and W. H. Schlesinger, editors. Year in Ecology and Conservation Biology 2009. Salvarina, I., E. Yohannes, B. M. Siemers, and K. Koselj. 2013. Advantages of using fecal samples for stable isotope analysis in bats: evidence from a triple isotopic experiment. Rapid Communications in Mass Spectrometry 27:1945-1953. Schoener, T. W. 1989. Food webs from the small to the large. Ecology 70:1559-1589. Shakesby, R. A. and S. H. Doerr. 2006. Wildfire as a hydrological and geomorphological agent. Earth-Science Reviews 74:269-307. Speir, S. L., M. M. Chumchal, R. W. Drenner, W. G. Cocke, M. E. Lewis, and H. J. Whitt. 2014. Methyl mercury and stable isotopes of nitrogen reveal that a

263

terrestrial spider has a diet of emergent aquatic insects. Environmental Toxicology and Chemistry 33:2506-2509. Spencer, C. N., K. O. Gabel, and F. R. Hauer. 2003. Wildfire effects on stream food webs and nutrient dynamics in Glacier National Park, USA. Forest Ecology and Management 178:141-153. Steinmetz, J., S. L. Kohler, and D. A. Soluk. 2003. Birds are overlooked top predators in aquatic food webs. Ecology 84:1324-1328. Suedel, B. C., J. A. Boraczek, R. K. Peddicord, P. A. Clifford, and T. M. Dillon. 1994. Trophic transfer and biomagnification potential of contaminants in aquatic ecosystems. Reviews of Environmental Contamination and Toxicology 136:21- 89. Sullivan, S. M. P. 2012. Geomorphic-ecological relationships highly variable between headwater and network mountain streams of northern Idaho, United States. Journal of the American Water Resources Association 1-12. Sullivan, S. M. P. and K. T. Vierling. 2012. Exploring the influences of multiscale environmental factors on the American dipper Cinclus mexicanus. Ecography 35. Sullivan, S. M. P. and M. C. Watzin. 2008. Relating stream physical habitat condition and concordance of biotic productivity across multiple taxa. Canadian Journal of Fisheries and Aquatic Sciences 65:2667-2677. Sullivan, S. M. P., M. C. Watzin, and W. C. Hession. 2004. Understanding stream geomorphic state in relation to ecological integrity: evidence using habitat assessments and macroinvertebrates. Environmental Management 34:669-673. Sullivan, S. M. P., M. C. Watzin, and W. S. Keeton. 2007. A riverscape perspective on habitat associations among riverine bird assemblages in the Lake Champlain Basin, USA. Landscape Ecology 22:1169-1186. Tagwireyi, T. and S. M. P. Sullivan. 2015. Distribution and trophic dynamics of riparian tetregnathid spiders in a large river system. Marine and Freshwater Research. Thompson, R. M. and C. R. Townsend. 1999. The effect of seasonal variation on the community structure and food-web attributes of two streams: implications for food-web science. Oikos 87:75-88. Van de Water, K. and M. North. 2010. Fire history of coniferous riparian forests in the Sierra Nevada. Forest Ecology and Management 260:384-395. Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Canadian Journal of Fishery and Aquatic Science 37:130-137. Vaughan, I. P., D. G. Noble, and S. J. Ormerod. 2007. Combining surveys of river habitats and river birds to appraise riverine hydromorphology. Freshwater Biology 52:2270-2284.

264

Verkaik, I., M. Rieradevall, S. D. Cooper, J. M. Melack, T. L. Dudley, and N. Prat. 2013a. Fire as a disturbance in mediterranean climate streams. Hydrobiologia 719:353-382. Verkaik, I., M. Vila-Escale, M. Rieradevall, and N. Prat. 2013b. Seasonal drought plays a stronger role than wildfire in shaping macroinvertebrate communities of Mediterranean streams. International Review of Hydrobiology 98:271-283. VTDEC. 2003. Stream geomorphic assessment handbook: rapid stream assessment - phase 2 field protocols. Vermont Agency of Natural Resources, Department of Environmental Conservation, Water Quality Division, Waterbury, VT, USA. Wallace, J. B., S. L. Eggert, J. L. Meyer, and J. R. Webster. 1997. Multiple trophic levels of a forested stream linked to terrestrial litter inputs. Science 277:102-104. Warfe, D. M., T. D. Jardine, N. E. Pettit, S. K. Hamilton, B. J. Pusey, S. E. Bunn, and P. M. Davies. 2013. Productivity, disturbance and ecosystem size have no influence on food chain length in seasonally connected rivers. PLoS ONE 10.1371/journal.pone.0066240. Willson, M. F. and K. M. Hocker. 2008. Natural history of nesting American dippers (Cinclus mexicanus) in southeastern Alaska. Northwestern Naturalist 89:97-106. Wondzell, S. M. and J. G. King. 2003. Postfire erosional processes in the Pacific Northwest and Rocky Mountain regions. Forest Ecology and Management 178:75-87.

265

Principle component “Fire axis” Independent variable loading r2 Fire frequency 1 0.83 5.71 2 0.90 6.74 3 0.74 4.56 4 0.42 1.48 > 4 0.70 4.02 Fire timing 5 years 0.80 5.32 10 years 0.90 6.75 20 years 0.88 6.38 30 years 0.97 7.76 Fire severity low 0.82 5.52 moderate 0.96 7.65 high 0.91 6.81

Eigenvalue 8.29 Percent variance explained 69.07

266

Table 14. Description of dipper breeding territories in Yosemite National Park, California, USA sampled in 2012 and 2013.

Catchments are Merced River, Tuolumne River, and South Fork of the Merced River. Drainage position indicates whether a dipper

territory is on a tributary, the mainstem, or at a confluence of a tributary and the Merced, Tuolumne, or South Fork of the Merced

Rivers. Stream order is based on Strahler (1952).

Dipper breeding Catchment Position Stream Headwater or Elevation RHA Condition territory order network (m) score (departure) Sampled in 2012

7 Bridalveil Creek Merced tributary 2 headwater 1997.4 133 Good (minor)

Bunnel Cascade Merced mainstem 4 network 1907.1 146 Reference (none) Covered Bridge South Fork Merced mainstem 4 network 1174.1 121 Good (minor) El Capitan Merced mainstem 5 network 1187.8 123 Good (minor) Superintendent's Bridge Merced mainstem 5 network 1201.8 146 Reference (none) Tenaya Creek Merced tributary 4 network 1280.2 128 Good (minor) Yosemite Falls Merced tributary 3 network 1212.2 140 Reference (none)

b Continued 267

Table14 continued

Sampled in 2013 Cathedral Creek Tuolumne confluence 5 network 1697.4 149 Reference (none) Clark Creek Merced tributary 2 headwater 2138.2 150 Reference (none) Illilouette Falls Merced confluence 4 network 1263.7 142 Reference (none) Morrisson Creek Tuolumne tributary 1 headwater 1724.3 153 Reference (none) Piute Creek Tuolumne tributary 3 network 1568.8 137 Reference (none) Register Creek Tuolumne confluence 5 network 1587.1 155 Reference (none) Regulation Creek Tuolumne confluence 5 network 1881.2 144 Reference (none)

6 Rodger's Creek Tuolumne mainstem 5 network 1476.5 151 Reference (none)

8 Washburn Lake Merced mainstem 4 network 2306.1 153 Reference (none) Sampled in 2012 and 2013 Bridalveil Falls Merced tributary 2 headwater 1224.7 136 Reference (none) Cascade Creek Merced tributary 2 headwater 1796.2 136 Reference (none) Cascade Falls Merced tributary 2 headwater 1074.7 114 Good (minor) Chilnualna Creek South Fork Merced confluence 3 network 1344.2 149 Reference (none) Illilouette Creek (a) Merced tributary 3 network 1922.1 140 Reference (none) Illilouette Creek (b) Merced tributary 3 network 1862.9 137 Reference (none)

b 268 Continued

Table 14 continued

Middle Tuolumne Tuolumne tributary 3 network 1773.6 128 Good (minor) Red Peak Merced confluence 4 network 2373.5 140 Reference (none) Snow Creek Merced tributary 4 network 1268.0 112 Good (minor) Tamarack Creek Merced tributary 1 headwater 1705.4 143 Reference (none) White Wolf Tuolumne tributary 2 headwater 2344.2 135 Good (minor)

b 269

Table 15. Descriptive statistics of independent and dependent variables used to assess the influence of wildfire, precipitation, and

ecosystem size (drainage area and territory length) on dipper reliance on aquatically-derived energy and trophic position. Fire

frequency is the proportion of each catchment draining to a dipper territory that has been burned 1, 2, 3, 4, or > 4 times since 1930.

Fire severity is the proportion of each catchment burned with low (dNBR = 1-2), moderate (dNBR = 3), or high (dNBR ≥ 4)

severity since 1984. Severity of overlapping fires were summed for each catchment. Fire timing is the proportion of each catchment

burned in the last 5, 10, 20, or 30 years. N = 27 for all variables except Hg (n = 18).

7 Minimum Median Maximum SD

0 Independent variables Drainage area (km2) 0.9 12.6 83.1 24.6 26.1 Precipitation (cm mo-1) 1.93 3.64 6.08 4.19 1.62 Territory length (m) 246 591 1664 744 397

Fire variables Fire frequency (%) 1 0.1 12.0 51.7 16.4 15.6 2 0.0 0.5 41.0 6.5 10.3 3 0.0 0.1 5.8 1.0 1.7 4 0.0 0.0 0.4 0.1 0.1

b 270 Continued

Table 15 continued

>4 0.0 0.0 0.0 0.0 0.0 Fire severity (%) Low 0.0 6.3 73.5 19.0 22.4 Moderate 0.0 0.4 14.6 2.7 4.0 High 0.0 0.1 14.4 1.6 3.2 Fire timing (%) 5 years 0.0 0.1 40.7 3.6 9.8 10 years 0.0 0.9 43.2 6.5 10.8 20 years 0.0 7.5 61.9 14.4 16.9 30 years 0.0 10.6 91.8 21.1 24.0

1 Dependent variables

Reliance on aquatically-derived energy 0.33 0.56 0.82 0.5 7 0.1 4 Trophic position 4.19 4.47 4.96 4.46 0.16

271

Table 16. Retained regression models (ΔAICc ≤ 4) with corresponding AICc scores, Akaike weights (wi), and variation explained

(R2). Null models (i.e., intercept only) are also included. Drainage area is in km2 and territory length is in m. Fire is the first axis of

principal component analysis describing fire severity, fire frequency, and fire timing. Precipitation is average monthly precipitation

for the water year corresponding to the sampling year.

2 American dipper response AIC AICc wi R Reliance on aquatically-derived energy (%) Drainage area (+) -33.65 0.00 0.43 0.48

2 Drainage area (+), precipitation (-) -32.56 1.09 0.25 0.52

Drainage area (+), fire (+) -31.33 2.32 0.14 0.49 Drainage area (+), territory length (+) -30.77 2.88 0.10 0.48 Drainage area (+), fire (+), precipitation (-) -30.19 3.46 0.08 0.53 Null -0.67 12.97 0.00 -

Continued

272

Table 16 continued

Trophic position Territory length (+) -18.07 0.00 0.30 0.13 Null -17.34 0.73 0.21 - Territory length (+), drainage area (+) -16.54 1.53 0.14 0.18 Territory length (+), precipitation (-) -16.13 1.94 0.11 0.18 Territory length (+), fire (-) -15.25 2.82 0.07 0.13 Drainage area (+) -14.95 3.12 0.06 0.01 Fire (-) -14.72 3.35 0.06 0.00

7 Precipitation (-) -14.71 3.36 0.06 0.00

b 273

Figure 25. Dipper breeding territories sampled in 2012 and 2013 in Yosemite National

Park, California, USA. Fire frequency is included to illustrate the differences in fire history among catchments draining to dipper breeding territories.

274

RHA Score

Figure 26. The negative relationship between RHA score and dipper territory length was significant (R2 = 0.17, F = 5.41, p = 0.045; y = 3351.0 – 18.51x).

275

276

0.9 a 0.8

0.7

0.6

derived derived energy - 0.5

0.4

0.3

0.2 headwater network 0.1 R2 = 0.07, p = 0.419 R2 = 0.16, p = 0.049

Reliance on aquatically on Reliance 0 0 2 4 6 8 10 12 Precipitation (cm month-1)

5.1 b headwater network 5 R2 = 0.00, p = 0.995 R2 = 0.03, p = 0.393 4.9

4.8

4.7

4.6

4.5

Trophic position Trophic 4.4

4.3

4.2

4.1 0 2 4 6 8 Precipitation (cm month-1)

Continued

277

Figure 27 continued

0.9 c 0.8

0.7

0.6

derived derived energy - 0.5

0.4

0.3

0.2 headwater network 0.1 R2 = 0.21, p = 0.251 R2 = 0.05, p = 0.431

Reliance on aquatically on Reliance 0 0 500 1000 1500 2000 Territory length (m)

5.1 d headwater network 5 R2 = 0.04, p = 0.655 R2 = 0.18, p = 0.101 4.9

4.8

4.7

4.6

4.5

Trophic position Trophic 4.4

4.3

4.2

4.1 0 500 1000 1500 2000 Territory length (m)

Continued

278

Figure 27 continued

0.9 e 0.8

0.7

0.6

derived derived energy - 0.5

0.4

0.3

0.2 headwater network 0.1 R2 = 0.66, p = 0.001 R2 = 0.01, p = 0.582

Reliance on aquatically on Reliance 0 -5 0 5 10 15 Fire axis

5.1 f headwater network 5 R2 = 0.03, p = 0.569 R2 = 0.00, p = 0.978 4.9

4.8

4.7

4.6

4.5

Trophic position Trophic 4.4

4.3

4.2

4.1 -5 0 5 10 15 Fire axis

279

Appendum B: Paired comparison of tree swallow blood and feces for stable isotope analysis

Methods

Study system

Riparian swallows were captured and sampled at nest boxes situated along the

Scioto River in central Ohio, USA. The Scioto River drains 16,882 km2, flowing through agricultural, forested, and urban landscapes. We collected blood and fecal samples from

17 individual adult and juvenile tree swallows (Tachycineta bicolor) from May to July

(2014) during the swallow reproductive season. Swallows were trapped at established nest boxes that were erected following design protocols set by the Golondrina de las

Américas (2011) project at Cornell University. Nest boxes were equipped with an external trapping mechanism which allows capture of adult swallows while inside of the nest box. We drew blood from a jugular vein following Sullivan and Vierling (2012). We immediately transferred blood samples to centrifuge tubes following collection. Blood samples were stored in 70% ethanol until analysis. Fecal samples were frozen until analysis.

In the laboratory, we dried feces samples for 48 hours at 60°C, then homogenized each sample with a mortar and pestle before packing in tin capsules for stable isotope analysis. Blood samples were dried for 48 hours in a 60° C oven then freeze dried using a

280

Labconco lyophilizer. Dried samples were pulverized (using a ceramic mortar and pestle) to ensure sample homogeneity and weighed (0.5-0.7 mg) and packed into 4 x 6 mm tin capsules. Continuous flow isotope-ratio mass spectrometry (EA-IRMS) was used to determine 13C and 15N for all samples at Washington State University Stable Isotope Core

(Pullman, Washington). The results are reported in δ (‰) notation defined as: δ13C or

15 13 12 15 14 δ N = [(Rsample/Rstandard)-1] * 100 where R is C/ C or N/ N, respectively. Typical analytical precision was 0.08‰ for δ15N and 0.19‰ for δ13C determination.

Data analysis

We used a two-sample t-test to determine if nutrient concentration (i.e., %C and

%N) and stable isotope ratios (i.e., δ13C and δ15N) were comparable for swallow blood and feces samples. Following comparison of means, we decided to explore the creation of a correction function for use of feces in place of blood in stable isotope analysis. To do this, we applied a simple linear regression to both δ13C and δ15N values for blood samples

15 versus feces samples. We applied the derived linear regression equations (δ N blood =

15 13 13 15 13 9.50 + 0.38 δ Nfeces and δ Cblood = -10.93 + 0.55 δ Cfeces; Figure 28) to δ N and δ C values of dipper feces in order to estimate values for dipper blood.

Results

We found no significant difference between blood and feces samples of riverine swallows for %C or δ13C (Table 17). However, %N was significantly lower in feces compared to blood and feces samples were significantly more depleted in δ15N compared to blood

281 samples (Table 17). Regression analysis revealed a significant linear relationship between blood and fecal samples for both δ13C and δ15N (Figure 28).

282

2014. Percent C and δ13C values did not differ between blood and feces, however %N was significantly lower and 15N significantly more depleted in feces samples compared to blood samples.

Blood Feces Response SD SD t p % C 48.63 11.00 41.53 15.11 1.61 0.137 δ13C -24.89 1.65 -25.64 2.92 -1.09 0.300 % N 14.06 3.35 9.67 4.12 -3.00 0.013 δ15N 11.78 1.69 7.25 3.06 -6.95 <0.0001

283

y = 9.50+ 0.38x

R2 = 0.37, p = 0.046

N of of blood(‰) N

15 δ

15 δ N of feces (‰)

y = -10.93+ 0.55x

R2 = 0.70, p = 0.001

C of blood (‰) of C

13 δ

13 δ C of feces (‰)

Figure 28. Linear regression of a) δ15N for blood and δ15N for feces; and b) δ13C for blood and δ13C for feces of riverine swallows collected from nest boxes along the Scioto

River in Columbus, OH in 2014. Dashed lines indicate 95% confidence curves.

284

Literature Cited

Golondrina de las Américas. 2011. Nest box design. Accessed March 24th, 2010. http://golondrinas.cornell.edu/Data_and_Protocol/NestBoxDesignCentimeters.ht ml. Sullivan, S. M. P. and K. T. Vierling. 2012. Exploring the influences of multiscale environmental factors on the American dipper Cinclus mexicanus. Ecography 35.

285

Conclusion

Wildfire has been shown to be a key disturbance to stream-riparian ecosystems, both creating and destroying habitat for in-stream and riparian organisms with implications for ecosystem function across space and time. Although I expected to find pronounced differences in the structure and function of stream-riparian ecosystems between stream reaches paired by fire severity and within catchments with a greater extent of frequenct, recent, or severe wildfire, the results of the four research chapters presented within this dissertation indicate that ecosystem size, aspects of climate (i.e., precipitation), and productivity/resource availability exerted greater influence on stream- riparian ecosystems than did fire in Yosemite National Park.

Most of the research relating fire to stream-riparian ecosystems has been conducted at the reach scale, focused on physical changes to ecosystem structure and populations of individual species, and has been located in temperate forests. I sought with this work to address some of the knowledge gaps arising from this historically-limited focus by several methods: 1) I utilized several aspects of wildfire as independent variables including severity, frequency, timing, and extent and derived these from both reach and catchment scale observations; 2) I measured both structural (i.e., riparian vegetation and stream geomorphology) and functional attributes (i.e., aquatic-terrestrial food-web connectivity and trophic position of top consumers) of stream-riparian ecosystems and their responses to wildfire; and 3) I used forested mountain streams

286 located within the Mediterranean climate region of California’s central Sierra Nevada as study locations. Results from the work presented in this dissertation support the hypothesis that in general ecosystem functioning of stream-riparian ecosystems may be resistant and resilient to terrestrial disturbances like wildfire. In addition, in ecosystems influenced by a Mediterranean-type climate characterized by annual dry-wet cycles and interannual variability in precipitation, climate and hydrology may play a much more ubiquitous role in determining stream-riparian ecosystem function that wildfire.

Collectiviely this work contributes to ecology theory as I found further empirical support for ecosystem size as a predictor of food-chain length (in this case trophic position of top consumers) and aquatic-terrestrial food web connectivity. Further, I found evidence that flood magnitude is a potential mechanism linking ecosystem size with food-web attributes. Coversely, terrestrial disturbance (i.e., wildfire) was relatively less supported as a driver of food-web dynamics in this system.

Future studies would benefit from continuing to focus on knowledge gaps within our understanding of wildfire effects on stream-riparian ecosystems. Specifically, future studies should take into account potential upstream-to-downstream effects of recent and historic wildfire and the potential importance of fire extent, frequency, and timing (not just severity) for driving stream-riparian ecosystem processes. In addition, the interaction between flow regimes, climate, and wildfire have just begun to receive attention, but are likely important drivers of stream-riparian ecosystem function and will be an important consideration of fire managers in a historic period of rapid climate change. Finally, studies that move away from measuring individual components of ecosystems (e.g.,

287 populations of individual species) to a more holistic analysis of ecosystem structure and function may provide a complimentary point of view toward the ultimate resistance and resilienace of stream-riparian ecosystems to wildfire.

288

Bibliography

Adams, D. E., R. C. Anderson, and S. L. Collins. 1982. Differential response of woody and herbaceous species to summer and winter burning in an Oklahoma grassland. The Southwest Naturalist 27:55-61. Agee, J. K. 1993. Fire ecology of the Pacific Northwest forests. Island Press, Washington, D.C. Agee, J. K., C. S. Wright, N. Williamson, and M. H. Huff. 2002. Foliar moisture content of Pacific Northwest vegetation and its relation to wildland fire behavior. Forest Ecology and Management 167:57-66. Alberts, J. M., S. M. P. Sullivan, and A. Kautza. 2013. Riparian swallows as integrators of landscape change in a multiuse river system: implications for aquatic-to- terrestrial transfers of contaminants. Science of the Total Environment 463:42-50. Allan, J. D. 1981. Determinants of diet of brook trout (Salvelinus fontinalis) in a mountain stream. Canadien Journal of Fishery and Aquatic Science 38:184-192. Allan, J. D., M. S. Wipfli, J. P. Caouette, A. Prussian, and J. Rogers. 2003. Influence of streamside vegetation on inputs of terrestrial invertebrates to salmonid food webs. Canadian Journal of Fisheries and Aquatic Sciences 60:309-320. Allen, C. D. and D. D. Breshears. 1998. Drought-induced shift of a forest-woodland ecotone: Rapid landscape response to climate variation. Proceedings of the National Academy of Sciences of the United States of America 95:14839-14842. Allen, D. C. and C. C. Vaughn. 2010. Complex hydraulic and substrate variables limit freshwater mussel species richness and abundance. Journal of the North American Benthological Society 29:383-394. Allred, B. W., S. D. Fuhlendorf, T. J. Hovick, R. D. Elmore, D. M. Engle, and A. Joern. 2013. Conservation implications of native and introduced ungulates in a changing climate. Global Change Biology 19:1875-1883. Alp, M., B. L. Peckarsky, S. M. Bernasconi, and C. T. Robinson. 2013. Shifts in isotopic signatures of animals with complex life-cycles can complicate conclusions on cross-boundary trophic links. Aquatic Sciences 75:595-606. Anderson, D. R. and K. P. Burnham. 2002. Avoiding pitfalls when using information- theoretic methods. Journal of Wildlife Management 66:912-918. Andreu, V., A. C. Imeson, and J. L. Rubio. 2001. Temporal changes in soil aggregates and water erosion after a wildfire in a Mediterranean pine forest. Catena 44:69-84. 289

Arbuckle, J. L. 2013. Amos (Version 22.0) SPSS, Chicago. Arim, M., P. A. Marquet, and F. M. Jaksic. 2007. On the relationship between productivity and food chain length at different ecological levels. American Naturalist 169:62-72. Arkle, R. S. and D. S. Pilliod. 2010. Prescribed fires as ecological surrogates for wildfires: A stream and riparian perspective. Forest Ecology and Management 259:893-903. Arkle, R. S., D. S. Pilliod, and K. Strickler. 2010. Fire, flow and dynamic equilibrium in stream macroinvertebrate communities. Freshwater Biology 55:299-314. Arnell, N. W. 2004. Climate change and global water resources: SRES emissions and socio-economic scenarios. Global Environmental Change-Human and Policy Dimensions 14:31-52. Bagne, K. E. and K. L. Purcell. 2011. Short-Term Responses of birds to prescribed fire in fire-suppressed forests of California. Journal of Wildlife Management 75:1051- 1060. Barbour, M. T., J. Gerritsen, B. D. Snyder, and J. B. Stribling. 1999. Rapid bioassessment protocols for use in streams and wadeable rivers: periphyton, benthic macroinvertebrates and fish. U.S. Environmental Protection Agency; Office of Water, Washington, D.C. Barro, S. C. and S. G. Conard. 1991. Fire effects on California chaparral systems - an overview. Environment International 17:135-149. Bartos, D. L. and R. B. Campbell Jr. 1998. Decline of quaking aspen in the interior west, examples from Utah. Rangelands 20:17-24. Baxter, C. V., K. D. Fausch, and W. C. Sounders. 2005. Tangled webs: reciprocal flows of invertebrate prey link streams and riparian zones. Freshwater Biology 50:20. Baxter, C. V., K. D. Fausch, M. Murikami, and P. L. Chapman. 2004. Fish invasion restructures stream and forest food webs by interrupting reciprocal prey subsidies. Ecology 85:2656-2663. Beakes, M. P., J. W. Moore, S. A. Hayes, and S. M. Sogard. 2014. Wildfire and the effects of shifting stream temperature on salmonids. Ecosphere 5. Bêche, L. A. and V. H. Resh. 2007. Short-term climatic trends affect the temporal variability of macroinvertebrates in California 'Mediterranean' streams. Freshwater Biology 52:2317-2339. Bêche, L. A., S. L. Stephens, and V. H. Resh. 2005. Effects of prescribed fire on a Sierra Nevada (California, USA) stream and its riparian zone. Forest Ecology and Management 218:37-59. Bellmore, J. R., C. V. Baxter, and P. J. Connolly. 2014. Spatial complexity reduces interaction strength in the meta-food web of a river floodplain mosaic. Ecology

290

Belwood, J. J. and M. B. Fenton. 1976. Variation in diet of Myotis lucifugus (Chiroptera- vespertilionidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie 54:1674-1678. Benda, L., D. Miller, P. Bigelow, and K. Andras. 2003. Effects of post-wildfire erosion on channel environments, Boise River, Idaho. Forest Ecology and Management 178:105-119. Benda, L., N. L. Poff, D. Miller, T. Dunne, G. Reeves, G. Pess, and M. Pollock. 2004. The network dynamic hypothesis: how channel networks structure riverine habitats. Bioscience 54:413-427. Bendix, J. and C. M. Cowell. 2010. Fire, floods and woody debris: Interactions between biotic and geomorphic processes. Geomorphology 116:297-304. Bertoldi, W., N. A. Drake, and A. M. Gurnell. 2011. Interactions between river flows and colonizing vegetation on a braided river: Exploring spatial and temporal dynamics in vegetation cover using satellite data. Earth Surface Processes and Landforms 36:1474-1486. Beschta, R. L., J. J. Rhodes, J. B. Kauffman, R. E. Gresswell, G. W. Minshall, J. R. Karr, D. A. Perry, F. R. Hauer, and C. A. Frissell. 2004. Postfire management on forested public lands of the Western United States. Conservation Biology 18:957- 969. Bess, E. C., R. R. Parmenter, S. McCoy, and M. C. Molles. 2002. Responses of a riparian forest-floor arthropod community to wildfire in the middle Rio Grande Valley, New Mexico. Environmental Entomology 31:774-784. Betts, E. F. and J. B. Jones, Jr. 2009. Impact of wildfire on stream nutrient chemistry and ecosystem metabolism in boreal forest catchments of interior Alaska. Arctic Antarctic and Alpine Research 41:407-417. Beyer, W. N., G. H. Heinz, and A. W. Redmon-Norwood. 1996. Environmental contaminants in wildlife: interpreting tissue concentrations. CRC Press, Boca Raton, FL. Bigelow, P. E., L. E. Benda, D. J. Miller, and K. M. Burnett. 2007. On debris flows, river networks, and the spatial structure of channel morphology. Forest Science 53:220- 238. Bird, M. I., E. Tait, C. M. Wurster, and R. W. Furness. 2008. Stable carbon and nitrogen isotope analysis of avian uric acid. Rapid Communications in Mass Spectrometry 22:3393-3400. Bisson, P.A., B.E., Rieman, C., Luce, P.F., Hessburg, D.C., Lee, J.L., Kersner, G.H., Reeves, and R.E., Gresswell. 2003. Fire and aquatic ecosystems of the Western USA: current knowledge and key questions. Forest Ecology and Management 178, 213–229.

291

Bladon, K. D., U. Silins, M. J. Wagner, M. Stone, M. B. Emelko, C. A. Mendoza, K. J. Devito, and S. Boon. 2008. Wildfire impacts on nitrogen concentration and production from headwater streams in southern Alberta's Rocky Mountains. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 38:2359-2371. Blais, J. M. 2005. Biogeochemistry of persistent bioaccumulative toxicants: processes affecting the transport of contaminants to remote areas. Canadian Journal of Fisheries and Aquatic Sciences 62:236-243. Blank, R. R., J. C. Chambers, and D. Zamudio. 2003. Restoring riparian corridors with fire: effects on soil and vegetation. Journal of Range Management 56:388-396. Blumenthal, S. A., K. L. Chritz, J. M. Rothman, and T. E. Cerling. 2012. Detecting intraannual dietary variability in wild mountain gorillas by stable isotope analysis of feces. Proceedings of the National Academy of Sciences of the United States of America 109:21277-21282. Boerner, R. E. J., J. Huang, and S. C. Hart. 2008. Impacts of fire and fire surrogate treatments on ecosystem nitrogen storage patterns: similarities and differences between forests of eastern and western North America. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 38:3056-3070. Bonada, N. and V. H. Resh. 2013. Mediterranean-climate streams and rivers: geographically separate but ecologically comparable freshwater systems. Hydrobiologia 719:1-29. Bozek, M. A. and M. K. Young. 1994. Fish mortality resulting from delayed-effects of fire in the greater Yellowstone ecosystem. Great Basin Naturalist 54:91-95. Brack, V. W. J. 1983. The nonhibernating ecology of bats in Indiana with emphasis on the endangered Indiana bat, Myotis sodalis. Purdue University, West Lafayette, Indiana, USA. Brigham, R. M., H. Aldridge, and R. L. Mackey. 1992. Variation in habitat use and prey selection by Yuma bats, Myotis yumanensis. Journal of Mammalogy 73:640-645. Brookshire, E. N. J. and K. A. Dwire. 2003. Controls on patterns of coarse organic particle retention in headwater streams. Journal of the North American Benthological Society 22:17-34. Buchalski, M. R., J. B. Fontaine, P. A. Heady, III, J. P. Hayes, and W. F. Frick. 2013. Bat response to differing fire severity in mixed-conifer forest California, USA. PLoS ONE 8. Buckland-Nicks, A., K. N. Hiller, and N. J. O'Driscoll. 2014. Mercury bioaccumulation in dragonflies (Odonata: Anisoptera): examination of life stages and body regions. Environmental Toxicology and Chemistry 2014. Buckton, S. T. and S. J. Ormerod. 2002. Global patterns of diversity among the specialist birds of riverine landscapes. Freshwater Biology 47:695-709. 292

Buhl, K. J. and S. J. Hamilton. 2000. Acute toxicity of fire-control chemicals, nitrogenous chemicals, and surfactants to rainbow trout. Transactions of the American Fisheries Society 129:408-418. Bull, W. B. 1979. Thresholds of critical power in streams. Geological Society of America Bulletin 90:453-464. Burdon, F. J. and J. S. Harding. 2008. The linkage between riparian predators and aquatic insects across a stream-resource spectrum. Freshwater Biology 53:330-346. Burnham, K. P. and D. R. Anderson. 2004. Multimodel inference - understanding AIC and BIC in model selection. Sociological Methods & Research 33:261-304. Burton, T. A. 2005. Fish and stream habitat risks from uncharacteristic wildfire: observations from 17 years of fire-related disturbances on the Boise National Forest, Idaho. Forest Ecology and Management 211:140-149. Carpenter, S. R., J. J. Cole, M. L. Pace, M. Van de Bogert, D. L. Bade, D. Bastviken, C. M. Gille, J. R. Hodgson, J. F. Kitchell, and E. S. Kritzberg. 2005. Ecosystem subsidies: terrestrial support of aqautic food webs from 13C addition to contrasting lakes. 86 10:2737-2750 Carrier, W. D. and B. Czech. 1996. Threatened and endangeres wildlife and livestock interactions.in P. R. Krausman, editor. Rangeland Wildlife. Society for Rangeleand Management, Denver, CO. Carrion, J. S., P. Sanchez-Gomez, J. F. Mota, R. Yll, and C. Chain. 2003. Holocene vegetation dynamics, fire and grazing in the Sierra de Gador, southern Spain. Holocene 13:839-849. Cerda, A. and S. H. Doerr. 2005. Influence of vegetation recovery on soil hydrology and erodibility following fire: an 11-year investigation. International Journal of Wildland Fire 14:423-437. Chen, C.-C. and Y. Wang. 2010. Relationships between stream habitat and breeding territory length of the Brown Dipper (Cinclus pallasii) in Taiwan. Journal of Ornithology 151:87-93. Cianfrani, C. M., S. M. P. Sullivan, W. C. Hession, and M. C. Watzin. 2009. Mixed stream channel morphologies: implications for fish community diversity. Aquatic Conservation: Marine and Freshwater Ecosystems 19:147-156. Coffman, G. C., R. F. Ambrose, and P. W. Rundel. 2010. Wildfire promotes dominance of invasive giant reed (Arundo donax) in riparian ecosystems. Biological Invasions 12:2723-2734. Cohen, J. E. and C. M. Newman. 1991. Community area and food chain length: theoretical predictions. American Naturalist 138:1542-1554. Collier, K. J., J. S. Bury, and M. Hibbs. 2002. A stable isotope study of linkages between stream and terrestrial food webs through spider predation. Freshwater Biology 47:1651-1659. 293

Coombs, J. S. and J. M. Melack. 2012. The initial impacts of a wildfire on hydrology and suspended sediment and nutrient export in California chaparral watersheds. Hydrological Processes. Cooper, S. D., H. M. Page, S. W. Wiseman, K. Klose, D. Bennet, T. Even, S. Sadro, C. E. Nelson, and T. L. Dudley. 2014. Physicochemical and biological responses of streams to wildfire severity in riparian zones. Freshwater Biology. e12523 Covich, A. P., M. A. Palmer, and T. A. Crowl. 1999. The role of benthic invertebrate species in freshwater ecosystems. BioScience 49:119-127. D'Antonio, C. M. and P. M. Vitousek. 1992. Biological invasions by exotic grasses, the grass/fire cycle, and global climate change. Annual Review of Ecology and Sytematics 23:63-89. Davies-Colley, R. J. and J. M. Quinn. 1998. Stream lighting in five regions of North Island, New Zealand: control by channel size and riparian vegetation. New Zealand Journal of Marine and Freshwater Research 32:591-605. Davis, F. W., E. A. Keller, A. Parikh, and J. Florsheim. 1989. Recovery of the chaparral riparian zone after wildfire. Pages 194-203 in D. L. Abell, editor. Proceedings of the California Riparian Systems Conference: Protection, Management, and Restoration for the 1990s. USDA General Technical Report Pacific Southwest, Davis, CA. Davis, J. C., G. W. Minshall, C. T. Robinson, and P. Landres. 2001. Monitoring wilderness stream ecosystems. USDA Forest Service, Ogden, UT. Davis, J. M., C. V. Baxter, E. J. Rosi-Marshall, J. L. Pierce, and B. T. Crosby. 2013. Anticipating stream ecosystem responses to climate change: toward predictions that incorporate effects via land-water linkages. Ecosystems 16:909-922. De Backer, A., C. Van Colen, M. Vincx, and S. Degraer. 2010. Complex hydraulic and substrate variables limit freshwater mussel species richness and abundance. Journal of the North American Benthological Society 29:383-384. DeBano, L. F. 2000. The role of fire and soil heating on water repellency in wildland environments: a review. Journal of Hydrology 231:195-206. DeFlorio, M. J., D. W. Pierce, D. R. Cayan, and A. J. Miller. 2013. Western US extreme precipitation events and their relation to ENSO and PDO in CCSM4. Journal of Climate 26:4231-4243. Doerr, S. H., A. J. D. Ferreira, R. P. D. Walsh, R. A. Shakesby, G. Leighton-Boyce, and C. O. A. Coelho. 2003. Soil water repellency as a potential parameter in rainfall- runoff modelling: experimental evidence at point to catchment scales from Portugal. Hydrological Processes 17:363-377. Downes, B. J., L. A. Marmuta, P. G. Fairweather, D. P. Faith, M. J. Keough, P. S. Lake, B. D. Mapstone, and G. P. Quinn. 2002. Monitoring ecological impacts: concepts and practice in flowing waters. Cambridge University Press, Cambridge, UK. 294

Dunham, J. B., A. E. Rosenberger, C. H. Luce, and B. E. Rieman. 2007. Influences of wildfire and channel reorganization on spatial and temporal variation in stream temperature and distribution of fish and amphibians. Ecosystems 10:335-346. Dunham, J. B., M. K. Young, R. E. Gresswell, and B. E. Rieman. 2003. Effects of fire on fish populations: landscape perspectives on peristance of native fishes and nonnative fish invasion. Forest Ecology and Management 178:183-196. Dwire, K. A. and J. B. Kauffman. 2003. Fire and riparian ecosystems in landscapes of the western USA. Forest Ecology and Management 178:61-74. Ealey, D. M. 1977. Aspects of the ecology and behavior of a breeding population of dippers (Cinclus mexicanus: Passeriformes) in southern Alberta., Edmonton, Alberta, CA. Earl, S. R. and D. W. Blinn. 2003. Effects of wildfire ash on water chemistry and biota in South-Western USA streams. Freshwater Biology 48:1015-1030. Eguchi, K. 1988. The ecological significant of time budgets of the brown dipper Cinclus- pallasii temminck. Japanese Journal of Ecology 38:243-256. Everett, R., R. Schellhaas, P. Ohlson, D. Spurbeck, and D. Keenum. 2003. Continuity in fire disturbance between riparian and adjacent sideslope Douglas-fir forest. Forest Ecology and Management 175:31-48. Fausch, K. D., C. E. Torgersen, C. V. Baxter, and H. W. Li. 2002. Landscapes to riverscapes: bridging the gap between research and conservation of stream fishes. Bioscience 52:483-498. Fausch, K. D., S. Nakano, and S. Kitano. 1997. Experimentally induced foraging mode shift by sympatric charrs in a Japanese mountain stream. Behavioral Ecology 8:414-420. Fausch, K. D., Y. Taniguchi, S. Nakano, G. D. Grossman, and C. R. Townsend. 2001. Flood disturbance regimes influence rainbow trout invasion success among five holarctic regions. Ecological Applications 11:1438-1455. Feck, J. and R. O. Hall Jr. 2004. Response of American dippers (Cinclus mexicanus) to variation in stream water quality. Freshwater Biology 49:1123-1137. Feio, M. J., C. N. Coimbra, M. A. S. Graca, S. J. Nichols, and R. H. Norris. 2010. The influence of extreme climatic events and human disturbance on macroinvertebrate community patterns of a Mediterranean stream over 15 years. Journal of the North American Benthological Society 29:1397-1409. Filipe, A. F., J. E. Lawrence, and N. Bonada. 2013. Vulnerability of stream biota to climate change in mediterranean climate regions: a synthesis of ecological responses and conservation challenges. Hydrobiologia 719:331-351. Finlay, J. C. 2001. Stable-carbon-isotope ratios of river biota, implications for energy flow in lotic food webs. Ecology 82:1052-1064.

295

Flannigan, M., B. Stocks, M. Turetsky, and M. Wotton. 2009. Impacts of climate change on fire activity and fire management in the circumboreal forest. Global Change Biology 15:549-560. Fleming, H. L., J. C. Jones, J. L. Belant, and D. M. Richardson. 2013. Multi-scale roost site selection by Rafinesque's Big-eared Bat (Corynorhinus rafinesquii) and Southeastern Myotis (Myotis austroriparius) in Mississippi. American Midland Naturalist 169:43-55. Flores, M., C. Kvamme, B. Rust, K. Takenaka, and D. Young. 2013. Rim Fire BAER - Soils Report. USDA Forest Service. Freeman, M. C., C. M. Pringle, and C. R. Jackson. 2007. Hydrologic connectivity and the contribution of stream headwaters to ecological integrity at regional scales. Journal of the American Water Resources Association 43:5-14. Frissell, C. A., W. J. Liss, C. E. Warren, and M. D. Hurley. 1986. A hierarchical framework for stream habitat classification: viewing streams in a watershed context. Environmental Management 10:199-215. Fukui, D., M. Murakami, S. Nakano, and T. Aoi. 2006. Effect of emergent aquatic insects on bat foraging in a riparian forest. Journal of Animal Ecology 75:1252-1258. Gaikowski, M. P., S. J. Hamilton, K. J. Buhl, S. F. McDonald, and C. H. Summers. 1996. Acute toxicity of three fire-retardant and two fire-suppressant foam formulations to the early life stages of rainbow trout (Oncorhynchus mykiss). Environmental Toxicology and Chemistry 15:1365-1374. Gamradt, S. C. and L. B. Kats. 1997. Impact of chaparral wildfire-induced sedimentation on oviposition of stream-breeding California newts (Taricha torosa). Oecologia 110:546-549. Gillespie, R. G. 1987. The mechanism of habitat selection in the long-jawed orb-weaving spider Tetragnatha elongata (Araneae, Tetragnathidae). Journal of Arachnology 15:81-90. Gimenez, A., E. Pastor, L. Zarate, E. Planas, and J. Arnaldos. 2004. Long-term forest fire retardants: a review of quality, effectiveness, application and environmental considerations. International Journal of Wildland Fire 13:1-15. Goldsmith, F. B., C. M. Harrison, and A. J. Morton. 1986. Description and analysis of vegetation.in P. D. Moore and S. B. Chapman, editors. Methods in Plant Ecology. Wiley & Sons, New York. Golondrina de las Américas. 2011. Nest box design. Accessed March 24th, 2010. http://golondrinas.cornell.edu/Data_and_Protocol/NestBoxDesignCentimeters.ht ml. Gotvald, A. J., N. A. Barth, A. G. Veilleux, and C. Parrett. 2012. Methods for determining magnitude and frequency of floods in California, based on data

296

through water year 2006. available online only at http://pubs.usgs.gov/sir/2012/5113/. Grafe, T. U., S. Dobler, and K. E. Linsenmair. 2002. Frogs flee from the sound of fire. Proceedings of the Royal Society - Biological Sciences 269:999-1003. Gregory, S. V., A. M. Gurnell, and K. L. Boyer. 2003. The ecology and management of wood in world rivers. American Fisheries Society, Bethesda, Maryland. Gregory, S. V., F. J. Swanson, W. A. McKee, and K. W. Cummins. 1991. An ecosystem perspective of riparian zones; focus on links between land and water. Bioscience 41:540-552. Gresswell, R. E. 1999. Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193-221. Grove, A.T. and Rackham, O. 2001. The nature of Mediterranean Europe: an ecological history. Journal of Mediterranean Ecology 3: 65-68. Gurnell, A. M., H. Piegay, F. J. Swanson, and S. V. Gregory. 2002. Large wood and fluvial processes. Freshwater Biology 47:601-619. Hagen, E. M. and J. L. Sabo. 2012. Influence of river drying and insect availability on bat activity along the San Pedro River, Arizona (USA). Journal of Arid Environments 84:1-8. Hall, S. J. and D. Lombardozzi. 2008. Short-term effects of wildfire on montane stream ecosystems in the Southern Rocky Mountains: one and two years post-burn. Western North American Naturalist 68:453-462. Halsey, R. W. 2008. Fire, chaparral, and survival in southern California. Sunbelt Publications, San Diego. Harris, H. E., C. V. Baxter, and J. M. Davis. In press. Debris flows mediate effects of wildfire on magnitude and composition of tributary subsidies to main-stem habitats. Freshwater Science. Harris, H. E., C. V. Baxter, and J. M. Davis. In revision. Wildfire and debris flows affect prey subsidies with implications for riparian and reverine predators. Freshwater Science. Hashiguchi, D. and S. Yamagishi. 1981. Dispersion pattern and territorial behavior of the wintering brown dipper Cinclus-pallasiiI. Japanese Journal of Ecology 31:161- 170. Hauer, F. R. and C. N. Spencer. 1998. Phosphorus and nitrogen dynamics in streams associated with wildfire: a study of immediate and longterm effects. International Journal of Wildland Fire 8:183-198. Henschel, J. R., D. Mahsberg, and H. Stumpf. 2001. Allochthonous aquatic insects increase predation and decrease herbivory in river shore food webs. Oikos 93:429-438.

297

Hickman, J. C., editor. 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. Hicks, J. H., M. S. Wipfli, D. W. Lang, and M. E. Lang. 2005. Marine-derived nitrogen and carbon in freshwater-riparian food webs of the Copper River Delta, southcentral Alaska. Oecologia 144:558-569. Hitt, N. P. 2003. Immediate effects of wildfire on stream temperature. Journal of Freshwater Ecology 18:171-173. Holt, R. D. 2004. Implications of system openess for local community structure and ecosystem function.in G. A. Polis, M. E. Power, and G. R. Huxel, editors. Food webs at the landscape level. The University of Chicago Press, Chicago, IL, U.S.A. Howell, P. J. 2006. Effects of wildfire and subsequent hydrologic events on fish distribution and abundance in tributaries of north fork John Day River. North American Journal of Fisheries Management 26:983-994. Hoyle, R. H. 1995. Structural equation modeling: concepts, issues, and applications. Sage Publications Inc., Thousand Oaks, CA. Hu, B., L. Yang, B. Wang, and L. Shan. 2005. Functional feeding groups of macroinvertebrates in 1 - 6 order tributaries of the Changjiang River. Chinese Journal of Applied and Environmental Biology 11:463-466. Hutto, R. L. 2008. The ecological importance of severe wildfire: some like it hot. Ecological Applications 18:1827-1834. Hynes, H. B. N. 1975. The stream and its valley. Verh. Internat. Verein. Limnol. 19:1-15. Isaak, D. J., C. H. Luce, B. E. Rieman, D. E. Nagel, E. E. Peterson, D. L. Horan, S. Parkes, and G. L. Chandler. 2010. Effects of climate change and wildfire on stream temperatures and salmonid thermal habitat in a mountain river network. Ecological Applications 20:1350-1371. Isaak, D. J., C. H. Luce, B. E. Rieman, D. E. Nagel, E. E. Peterson, D. L. Horan, S. Parkes, and G. L. Chandler. 2010. Effects of climate change and wildfire on stream temperatures and salmonid thermal habitat in a mountain river network. Ecological Applications 20:1350-1371. Jackson, B. K. and S. M. P. Sullivan. 2009. Influence of fire severity on riparian vegetation heterogeneity in an Idaho, U.S.A. wilderness. Forest Ecology and Management 259:24-32. Jackson, B. K. and S. M. P. Sullivan. In press. Responses of riparian tetragnathid spiders to wildfire in forested ecosystems of the California Mediterranean climate region, USA Freshwater Science. Jackson, B. K., S. M. P. Sullivan, and R. L. Malison. 2012. Wildfire severity mediates fluxes of plant material and terrestrial invertebrates to mountain streams. Forest Ecology and Management 278:27-34.

298

Jackson, B. K., S. M. P. Sullivan, C. V. Baxter, and R. L. Malison. In press. Stream- riparian ecosystems.in D. A. Dellasala and C. T. Hanson, editors. Benefits of high-severity fire for biodiversity and conservation. Elsevier. Jardine, T. D., B. J. Pusey, S. K. Hamilton, N. E. Pettit, and P. M. Davies. 2012. Fish mediate high food web connectvity in the lower reaches of a tropical floodplain river. Oecologia 168:829-838. Johansson, M. E. and C. Nilsson. 2002. Responses of riparian plants to flooding in free flowing and regulated rivers: an experimental study. Journal of Applied Ecology 39:917-986. Johnson, J.B. and K.S. Ormland. 2004. Model selection in ecology and evolution. Trends in Ecology and Evolution 19:101-108. Jones, J. R. and N. V. DeByle. 1985. Fire. Pages 77-81 in N. V. DeByle and R. P. Winokur, editors. Aspen: ecology and management in the western United States. USDA Forest Service, Ogden, UT. Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood pulse concept in river- floodplain systems. Pages 110-127 in International Large River Symposium. Karr, J. R., J. J. Rhodes, G. W. Minshall, F. R. Hauer, R. L. Beschta, C. A. Frissell, and D. A. Perry. 2004. The effects of postfire salvage logging on aquatic ecosystems in the American West. Bioscience 54:1029-1033. Kato, C., T. Iwata, S. Nakano, and D. Kishi. 2003. Dynamics of aquatic insect flux affects distribution of riparian web-building spiders. Oikos 103:113-120. Kauffman, J. B. and R. E. Martin. 1989. Fire behavior, fuel consumption, and forest-floor changes following prescribed understory fires in Sierra-Nevada mixed conifer forests. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 19:455-462. Kauffman, J. B. and R. E. Martin. 1990. Sprouting shrub response to different seasons and fuel consumtion levels of prescribed fire in Sierra-Nevada mixed conifer ecosystems. Forest Science 36:748-764. Keeley, J. E., H. Safford, C. J. Fotheringham, J. Franklin, and M. Moritz. 2009. The 2007 southern California wildfires: lessons in complexity. Journal of Forestry 107:287- 296. Kennedy, T. L. and T. F. Turner. 2011. River channelization reduces nutrient flow and macroinvertebrate diversity at the aquatic terrestrial transition zone. Ecosphere 2:1-13. Key, C.H. and N.C. Benson. 2006. Landscape assessment: ground measure of severity, the Composite Burn Index; and remote sensing of severity, the Normalized Burn Ratio. In D.C. Lutes; R.E. Keane; J.F. Caratti; C.H. Key; N.C. Benson; S. Sutherland; and L.J. Gangi. 2006. FIREMON: Fire effects monitoring and

299

inventory system. USDA Forest Service, Rocky Mountain Research Station, Ogden, UT. Gen. Tech. Rep. RMRS-GTR-164-CD: LA 1-51. Keyser, T. L., F. W. Smith, and W. D. Shepperd. 2005. Trembling aspen response to a mixed-severity wildfire in the Black Hills, South Dakota, USA. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 35:2679-2684. Khorram, S. and E. F. Katibah. 1984. Vegetation and land cover mapping of the middle fork of the Feather River watershed using landsat digital data. Forest Science 30:248-258. Kidd, K. A., D. W. Schindler, R. H. Hesslein, and D. C. G. Muir. 1995. Correlation between stable nitrogen isotope ratios and concentrations of organochlorides in biota from a fresh-water food-web. Science of the Total Environment 160-61:381- 390. Kingery, H. E. 1996. American Dipper (Cinclus mexicanus).in A. Poole and F. Gill, editors. The Birds of North America. The Academy of Natural Sciences and The American Ornithologists Union, Philadelphia, P.A. and Washington, D.C. Kirkpatrick, C. and C. J. Conway. 2010. Importance of montane riparian forest and influence of wildfire on nest-site selection of ground-nesting birds. Journal of Wildlife Management 74:729-738. Kobziar, L. N. and J. R. McBride. 2006. Wildfire burn patterns and riparian vegetation response along two Sierra Nevada streams. Forest Ecology and Management 222:254-265. Koetsier, P., Q. Tuckett, and J. White. 2007. Present effects of past wildfires on the diets of stream fish. Western North American Naturalist 67:429-438. Koetsier, P., T. R. B. Krause, and Q. M. Tuckett. 2010. Present effects of past wildfires on leaf litter breakdown in stream ecosystems. Western North American Naturalist 70:164-174. Kotliar, N. B., S. J. Hejl, R. L. Hutto, V. A. Saab, C. P. Melcher, and M. E. McFadzen. 2002. Effects of fire and post-fire salvage logging on avian communities in conifer-dominated forests of the western United States. Studies in Avian Biology. 25, 49-64. Koyama, A., K. L. Kavanagh, and K. Stephan. 2010. Wildfire effects on soil gross nitrogen transformation rates in coniferous forests of central Idaho, USA. Ecosystems 13:1112-1126. Kruse, A. and J. Piehl. 1986. The impact of prescribed burning on groud-nesting birds. in G. Clambey and R. Pemble, editors. The prairie: past present and future. Proceedings of the 9th North American Prairie Conference, Moorhead, MN, Center for Environmental Studies, Tri-College University, Fargo, North Dakota, USA.

300

Lane, E. W. 1955. The importance of fluvial morphology in hydraulic engineering. Proceedings of the American Society of Civil Engineers 745:1-17. Lavabre, J., D. S. Torres, and F. Cernesson. 1993. Changes in the hydrological response of a small Mediterranean basin a year after a wildfire. Journal of Hydrology 142:273-299. Lawrence, J. E., K. B. Lunde, R. D. Mazor, L. A. Beche, E. P. McElravy, and V. H. Resh. 2010. Long-term macroinvertebrate responses to climate change: implications for biological assessment in mediterranean-climate streams. Journal of the North American Benthological Society 29:1424-1440. Legleiter, C. J., R. L. Lawrence, M. A. Fonstad, W. A. Marcus, and R. Aspinall. 2003. Fluvial response a decade after wildfire in the northern Yellowstone ecosystem: a spatially explicit analysis. Geomorphology 54:119-136. Lenihan, J. M., R. Drapek, D. Bachelet, and R. P. Neilson. 2003. Climate change effects on vegetation distribution, carbon, and fire in California. Ecological Applications 13:1667-1681. Lewanzik, D., D. H. Kelm, S. Greiner, M. Dehnhard, and C. C. Voigt. 2012. Ecological correlates of cortisol levels in two bat species with contrasting feeding habits. General and Comparative Endocrinology 177:104-112. Linder, K. A. and S. H. Anderson. 1998. Nesting habitat of Lewis' Woodpeckers in southeastern Wyoming. Journal of Field Ornithology 69:109-116. Logie, J. W., D. M. Bryant, D. L. Howell, and J. A. Vickery. 1996. Biological significance of UK critical load exceedance estimates for flowing waters: Assessments of dipper Cinclus cinclus populations in Scotland. Journal of Applied Ecology 33:1065-1076. Mack, M. C. and C. M. D'Antonio. 1998. Impacts of biological invasions on disturbance regimes. Trends in Ecology and Evolution 13:195-198. Malison, R. L. and C. V. Baxter. 2010. The fire pulse: wildfire stimulates flux of aquatic prey to terrestrial habitats driving increases in riparian consumers. Canadian Journal of Fisheries and Aquatic Sciences 67:570-579. Malison, R. L. and C. V. Baxter. 2010. Effects of wildfire of varying severity on benthic stream insect assemblages and emergence. Journal of the North American Benthological Society 29:1324-1338. Malmon, D. V., S. L. Renean, D. Katzman, A. Lavine, and J. Lyman. 2007. Suspended sediment transport in an ephemeral stream following wildfire. Journal of Geophysics Research - Earth Surface 112:FO2006. Marks, J. C., M. E. Power, and M. S. Parker. 2000. Flood disturbance, algal productivity, and interannual variation in food chain length. Oikos 90:20-27. Marzolin, G. 2002. Influence of the mating system of the Eurasian dipper on sex-specific local survival rates. Journal of Wildlife Management 66:1023-1030. 301

Mast, M. A. and D. W. Clow. 2008. Effects of 2003 wildfires on stream chemistry in Glacier National Park, Montana. Hydrological Processes 22:5013-5023. May, C. 2007. Sediment and wood routing in steep headwater streams: an overview of geomorphic processes and their topographic signatures. Forest Science 53:119- 130. May, C. L. and R. E. Gresswell. 2003. Processes and rates of sediment and wood accumulation in headwater streams of the Oregon Coast Range, USA. Earth Surface Processes and Landforms 28:409-424. Mayor, A. G., S. Bautista, J. Llovet, and J. Bellot. 2007. Post-fire hydrological and erosional responses of a Mediterranean landscape: seven years of catchment-scale dynamics. Catena 71:68-75. McCann, K. S. and N. Rooney. 2009. The more food webs change, the more they stay the same. Philos T R Soc B 364. McCune, B., J. B. Grace, and D. L. Urban. 2002. Analysis of ecological communities. MjM Software Design, Gleneden Beach, OR. McCune, B., Mefford, M.J., 1999. PC-ORD: Multivariate Analysis of Ecological Data. MjM Software Design, Gleneden Beach, OR. McHugh, P. A., A. R. McIntosh, and P. G. Jellyman. 2010. Dual influences of ecosystem size and disturbance on food chain length in streams. Ecology Letters 13:881-890. McIver, J. D. and L. Starr. 2001. A literature review on the environmental effects of postfire logging. Western Journal of Applied Forestry 16:159-168. McIver, J. D. and R. McNeil. 2006. Soil disturbance and hill-slope sediment transport after logging of a severely burned site in northeastern Oregon. Western Journal of Applied Forestry 21:123-133. McIver, J. D., S. L. Stephens, J. K. Agee, J. Barbour, R. E. J. Boerner, C. B. Edminster, K. L. Erickson, K. L. Farris, C. J. Fettig, C. E. Fiedler, S. Haase, S. C. Hart, J. E. Keeley, E. E. Knapp, J. F. Lehmkuhl, J. J. Moghaddas, W. Otrosina, K. W. Outcalt, D. W. Schwilk, C. N. Skinner, T. A. Waldrop, C. P. Weatherspoon, D. A. Yaussy, A. Youngblood, and S. Zack. 2013. Ecological effects of alternative fuel- reduction treatments: highlights of the National Fire and Fire Surrogate study (FFS). International Journal of Wildland Fire 22:63-82. McMichael, C. E., A. S. Hope, D. A. Roberts, and M. R. Anaya. 2004. Post-fire recovery of leaf area index in California chaparral: a remote sensing-chronosequence approach. International Journal of Remote Sensing 25:4743-4760. Mellon, C. D., M. S. Wipfli, and J. L. Li. 2008. Effects of forest fire on headwater stream macroinvertebrate communities in eastern Washington, USA. Freshwater Biology 53:2331-2343.

302

Meyer, G. A. and J. L. Pierce. 2003. Climatic controls on fire-induced sediment pulses in Yellowstone National Park and central Idaho: a long-term perspective. Forest Ecology and Management 178:89-104. Meyer, J. L., D. L. Strayer, J. B. Wallace, S. L. Eggert, G. S. Helfman, and N. E. Leonard. 2007. The contribution of headwater streams to biodiversity in river networks. Journal of the American Water Resources Association 43:86-103. Meyer, L. A. and S. M. P. Sullivan. 2013. Bright lights, big city: influences of ecological light pollution on reciprocal stream-riparian invertebrate fluxes. Ecological Applications 23:1322-1330. Mihuc, T. B. and G. W. Minshall. 1995. Trophic generalists vs trophic specialists - implications for food-web dynamics in post-fire streams. Ecology 76:2361-2372. Mihuc, T. B. and G. W. Minshall. 2005. The trophic basis of reference and post-fire stream food webs 10 years after wildfire in Yellowstone National Park. Aquatic Sciences 67:541-548. Miller, C. and D. L. Urban. 1999. Forest pattern, fire, and climatic change in the Sierra Nevada. Ecosystems 2:76-87. Miller, D., C. Luce, and L. Benda. 2003. Time, space, and episodicity of physical disturbance in streams. Forest Ecology and Management 178:121-140. Miller, J. D., H. D. Safford, M. Crimmins, and A. E. Thode. 2009. Quantitative evidence for increasing forest fire severity in the Sierra Nevada and Southern Cascade mountains, California and Nevada, USA. Ecosystems 12:16-32. Miller, M. 2000. Fire autecology. Pages 9-34 in J.K. Brown and J.K. Smith, editors. Wildland fire and ecosystems: effects of fire on flora. General Technical Report RMRS-GTR-412-vol2, USDA Forest Service, Ogden, UT. Miller, R.G., 1981. Simultaneous Statistical Inference. Springer Verlag, pp. 6-8. Milly, P. C. D., K. A. Dunne, and A. V. Vecchia. 2005. Global pattern of trends in streamflow and water availability in a changing climate. Nature 438:347-350. Minshall, G. F., J. T. Brock, D. A. Andrews, and C. T. Robinson. 2001. Water quality, substratum and biotic responses of five central Idaho (USA) streams during the first year following the Mortar Creek fire. International Journal of Wildland Fire 10:185-199. Minshall, G. W. 1978. Autotrophy in stream ecosystems. Bioscience 28:767-770. Minshall, G. W. 2003. Responses of stream benthic macroinvertebrates to fire. Forest Ecology and Management 178:155-161. Minshall, G. W., C. T. Robinson, and D. E. Lawrence. 1997. Postfire responses of lotic ecosystems in Yellowstone National Park, USA. Canadian Journal of Fisheries and Aquatic Sciences 54:2509-2525.

303

Minshall, G. W., C. T. Robinson, D. E. Lawrence, D. A. Andrews, and J. T. Brock. 2001. Benthic macroinvertebrate assemblages in five central Idaho (USA) streams over a 10-year period following disturbance by wildfire. International Journal of Wildland Fire 10:201-213. Minshall, G. W., J. T. Brock, and J. D. Varley. 1989. Widlfire and Yellowstone's stream ecosystems. Bioscience 39:707-718. Minshall, G. W., K. E. Bowman, B. A. Rugenski, and C. Relyea. 2003. Monitoring of streams in the Payette National Forest 1988-2003: Big Creek and South Fork Salmon tributaries pre- and post-fire. USGS, Payette National Forest. Minshall, G. W., T. V. Royer, and C. T. Robinson, editors. 2004. Stream ecosystem responses following the Yellowstone wildfires: the first 10 years. Yale University Press, New Haven, CT. Montgomery, D. R. and J. M. Buffington. 1997. Channel-reach morphology in mountain drainage basins. GSA Bulletin 109:596-611. Moody, J. A. and D. A. Martin. 2001. Post-fire, rainfall intensity-peak discharge relations for three mountainous watersheds in the western USA. Hydrological Processes 15:2981-2993. Moritz, M. A., M.-A. Parisien, E. Batllori, M. A. Krawchuk, J. Van Dorn, D. J. Ganz, and K. Hayhoe. 2012. Climate change and disruptions to global fire activity. Ecosphere 3. Morrissey, C. A., L. I. Bendell-Young, and J. E. Elliot. 2004. Linking contaminant profiles to the diet and breeding locations of American dippers using stable isotopes. Journal of Applied Ecology 41:502-512. Muggeo, V M R. 2015. Segmented: Regression models with breakpoints/changepoints estimation. R package version 0.5-1.1. Available at: http://cran.r- project.org/web/packages/segmented/index.html. Muggeo, V.M.R. 2003. Estimating regression models with unknown break-points. Statistics in Medicine, 22: 3055-3071. Muller-Dombois, D. and H. Ellensburg. 1974. Aims and methods of vegetation ecology. Wiley & Sons, New York. Murakami, M. and S. Nakano. 2002. Indirect effect of aquatic insect emergence on a terrestrial insect population through bird predation. Ecology Letters 5:333. Naiman, R. J. and H. Decamps. 1997. The Ecology of Interfaces: Riparian Zones. Annual Review of Ecology and Systematics 28:621-650. Naiman, R. J., H. Decamps, and M. E. McClain. 2005. Riparia: Ecology, Conservation, and Management of Streamside Communities. Elsevier Academic Press, Burlington, MA.

304

Nakano, S., H. Miyasaka, and N. Kuhara. 1999. Terrestrial-aquatic linkages: riparian arthropod inputs alter trophic cascades in a stream food web. Ecology 80:2435- 2441. Neville, H., J. Dunham, A. Rosenberger, J. Umek, and B. Nelson. 2009. Influences of wildfire, habitat size, and connectivity on trout in headwater streams revealed by patterns of genetic diversity. Transactions of the American Fisheries Society 138:1314-1327. Ober, H. K. and J. P. Hayes. 2008. Influence of vegetation on bat use of riparian areas at multiple spatial scales. Journal of Wildlife Management 72:396-404. Ormerod, S. J. 1985. The diet of breeding dippers (Cinclus mexicanus) and their nestlings in the catchment of the River Wye, mid-Wales: a preliminary study by faecal analysis. Ibis 127:316-331. Ormerod, S. J., J. O. Halloran, S. D. Gribbin, and S. J. Tyler. 1991. The ecology of dippers Cinclus cinclus in relation to stream acidity in upland wales: breeding performance, calcium physiology and nestling growth. Journal of Applied Ecology 28:419-433. Ormerod, S. J., M. Dobson, A. G. Hildrew, and C. R. Townsend. 2010. Multiple stressors in freshwater ecosystems. Freshwater Biology 55:1-4. Paetzold, A. C., C. J. Schubert, and K. Tockner. 2005. Aquatic-terrestrial linkages along a braided river: riparian arthropods feeding on aquatic insects. Ecosystems 8:748- 758. Parker, M. S. 1994. Feeding ecology of stream-dwelling Pacific giant salamander larvae (Dicamptodon tenebrosus). Copeia 1994:705-718. Parker, S. M. and A. D. Huryn. 2006. Food web structure and function in two arctic streams with contrasting disturbance regimes. Freshwater Biology 51:1249-1263. Parnell, A. and A. Jackson. 2013. Stable Isotope Analysis in R. http://cran.r- project.org/web/pakage/siar/. Parnell, A. and A. Jackson. 2013. Stable isotope analysis in R. http://cran.r- project.org/web/pakage/siar/. Parnell, A., R. Inger, S. Bearhop, and A. Jackson. 2010. Coping with too much variation. PLoS ONE 5:e9672. Pausas, J. G. and S. Paula. 2012. Fuel shapes the fire-climate relationship: evidence from Mediterranean ecosystems. Global Ecology and Biogeography 21:1074-1082. Pettit, N. E. and R. J. Naiman. 2007. Fire in the riparian zone: characteristics and ecological consequences. Ecosystems 10:673-687. Pettit, N. E. and R. J. Naiman. 2007. Postfire response of flood-regenerating riparian vetation in a semi-arid landscape. Ecology 88:2094-2104.

305

Piccolo, J. J. and M. S. Wipfli. 2002. Does red alder (Alnus rubra) in upland riparian forests elevate macroinvertebrate and detritus export from headwater streams to downstream habitats in southeastern Alaska? Canadian Journal of Fisheries and Aquatic Sciences 59:503-513. Pilliod, D. S., R. B. Bury, E. J. Hyde, C. A. Pearl, and P. S. Corn. 2003. Fire and amphibians in North America. Forest Ecology and Management 178:163-181. Pimm, S. L. 2002. Food webs. The University of Chicago Press, Chicago. Pimm, S. L. and J. H. Lawton. 1977. Number of trophic levels in ecological communities. Nature 268:329-331. Podlesak, D. W., S. R. McWilliams, and K. A. Hatch. 2005. Stable isotopes in breath, blood, feces and feathers can indicate intra-individual changes in the diet of migratory songbirds. Oecologia 142:501-510. Poff, L. N. 1997. Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16:391-409. Poff, N. L. 1992. Why disturbances can be predictable: a prespective on the definition of disturbance in streams. Journal of the North American Benthological Society 11:86-92. Poff, N. L. and J. V. Ward. 1990. Physical habitat template of lotic systems - recovery in the context of historical pattern of spatiotemporal heterogeneity. Environmental Management 14:629-645. Poff, N. L., J. D. Olden, D. M. Merritt, and D. M. Pepin. 2007. Homogenization of regional river dynamics by dams and global biodiversity implications. Proceedings of the National Academy of Sciences of the United States of America 104:5732-5737. Polis, G. A., W. B. Anderson, and R. D. Holt. 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Review of Ecology and Systematics 28:289-316. Polvi, L. E. 2011. Geomorphic and process domain controls on riparian zones in the Colorado Front Range. Geomorphology 125:504. Post, D. M. 2002. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83:703-718. Post, D. M. 2002a. The long and short of food-chain length. Trends in Ecology & Evolution 17:269-278. Postel, S. and S. Carpenter. 1997. Freshwater ecosystem services. Pages 195-214 in G. Daily, editor. Nature's services: societal dependence on natural ecosystems. Island Press.

306

Power, M. E. 1992. Habitat heterogeneity and the functional significance of fish in river food webs. Ecology 73:1675-1688. Power, M. E. 2002. River-watershed exchange: effects of riverine subsidies on riparian lizards and their terrestrial prey. Ecology 83. Power, M. E. 2006. Environmental controls on food web regimes: a fluvial perspective. Progress in Oceanography 68:125-133. Power, M. E. and W. E. Dietrich. 2002. Food webs in river networks. Ecological Research 17:451-471. Power, M. E. and W. E. Rainey. 2000. Food webs and resource sheds: towards spatially delimiting trophic interactions. Pages 291-314 in M. J. Hutchings, E. A. John, and A. J. Stewart, editors. The ecological consequences of environmental heterogeneity. Blackwell Science, London, England. Power, M. E., M. S. Parker, and W. E. Dietrich. 2008. Seasonal reassembly of a river food web: floods, droughts, and impacts of fish. Ecological Monographs 78:263- 282. Power, M. E., W. E. Rainey, M. S. Parker, J. L. Sabo, A. Smyth, S. Khandwala, J. C. Finlay, F. C. McNeely, K. Marsee, and C. Anderson. 2004. River to watershed subsidies in an old-growth conifer forest. Pages 217-240 in G. A. Polis, M. E. Power, and G. R. Huxel, editors. Food webs at the landscape level. The University of Chicago Press, Chicago, IL, USA. Price, F. E. and C. E. Bock. 1983. Population ecology of the dipper (Cinclus mexicanus) in the Front Range of Colorado. Studies in Avian Biology:1-84. Propst, D. L., J. A. Stefferud, and P. R. Turner. 1992. Conservation and status of Gila trout, Oncorhynchus gilae. Southwestern Naturalist 37:117-125. Protasov, A. A. 2008. River and lake continua: an attempt at analysis and synthesis. Inland Water Biology 1:105-113. Pyne, S. J. 1997. America's fires: management on wildlands and forests. Forest History Society, Durham, North Carolina. Pyne, S. J. 2004. Tending Fire: Coping with America's Wildland Fires. Washington Island Press. Quigley, T. M. and S. J. Arbelbide. 1997. An assessment of ecosystem components in the Interior Columbia Basin and portions of the Klamath and Great Basins. General Technical Report PNW-GTR-405. USDA Forest Service, Portland, Oregon, USA. Rambo, T. R. and M. P. North. 2008. Spatial and temporal variability of canopy microclimate in a Sierra Nevada riparian forest. Northwest Science 82:259-268. Rehn, A. C. 2010. The effects of wildfire on benthic macroinvertebrates in southern California. San Diego, CA.

307

Resh, V. H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W. Li, G. W. Minshall, S. R. Reice, A. L. Sheldon, J. B. Wallace, and R. C. Wissmar. 1988. The role of disturbance in stream ecology. Journal of the North American Benthological Society 7:433-455. Rice, S., M. Stoffel, J. M. Turowski, and A. Wolf. 2012. Disturbance regimes at the interface of geomorphology and ecology. Earth Surface Processes and Landforms 37:1678-1682. Rieman, B. E. and J. Clayton. 1997. Wildfire and native fish: issues of forest health and conservation of sensitive species. Fisheries 22:6-15. Rieman, B. E., D. Lee, G. Chandler, and D. Meyer. 1997. Does wildfire threaten extinctions for salmonids? Responses or redband trout and bull trout following recent large fires on the Boise National Forest. Pages 47-57 in Symposium on fire effects on threatened and endangered species and habitats. International Association of Wildland Fire, Fairfield. Rinne, J. N. 1996. Short-term effects of wildfire on fishes and aquatic macroinvertebrates in the southwestern United States. North American Journal of Fisheries Management 16:653-658. Robinson, C. T., G. W. Minshall, and T. V. Royer. 2000. Inter-annual patterns in macroinvertebrate communities of wilderness streams in Idaho, USA. Hydrobiologia 421:187-198. Robinson, C. T., K. Tockner, and P. Burgherr. 2004. Drift benthos relationships in the seasonal colonization dynamics of alpine streams. Archiv Fur Hydrobiologie 160:447-470. Robinson, C. T., S. R. Rushforth, and G. W. Minshall. 1994. Diatom asssemblages of streams influenced by wildfire. Journal of Phycology 30:209-216. Robinson, C. T., U. Uehlinger, and G. W. Minshall. 2005. Functional characteristics of wilderness streams twenty years following wildfire. Western North American Naturalist 65:1-10. Robinson, C. T., U. Uehlinger, and M. T. Monaghan. 2003. Effects of a multi-year experimental flood regime on macroinvertebrates downstream of a reservoir. Aquatic Sciences 65:210-222. Romero, N., R. E. Gresswell, and J. L. Li. 2005. Changing patterns in coastal cutthroat trout (Oncorhynchus clarki clarki) diet and prey in a gradient of deciduous canopies. Canadian Journal of Fisheries and Aquatic Sciences 62:1797-1807. Romme, W. H., M. G. Turner, R. H. Gardner, W. W. Hargrove, G. A. Tuskan, D. G. Despain, and R. A. Renkin. 1997. A rare episode of sexual reproduction in Aspen (Populus tremuloides Michx) following the 1988 Yellowstone fires. Natural Areas Journal 17:17-25.

308

Romme, W. H., M. S. Boyce, R. Gresswell, E. H. Merrill, G. W. Minshall, C. Whitlock, and M. G. Turner. 2011. Twenty years after the 1988 Yellowstone fires: lessons about disturbance and ecosystems. Ecosystems 14:1196-1215. Rood, S. B., L. A. Goater, J. M. Mahoney, C. M. Pearce, and D. G. Smith. 2007. Floods, fire, and ice: disturbance ecology of riparian cottonwoods. Canadian Journal of Botany 85:1019-1032. Rosenberger, A. E., J. B. Dunham, J. M. Buffington, and M. S. Wipfli. 2011. Persistent effects of wildfire and debris flows on the invertebrate prey base of Rainbow Trout in Idaho streams. Northwest Science 85:55-63. Rosgen, D.L., 1996. Applied River Morphology. Wildland Hydrology, Pagosa Springs, CO. Rosi-Marshall, E. J. and J. B. Wallace. 2002. Invertebrate food webs along a stream resources gradient. Freshwater Biology 47:129-141. Rowse, L. M., A. D. Rodewald, and S. M. P. Sullivan. 2014. Pathways and consequences of contaminant flux to Acadian flycatchers (Empidonax virescens) in urbanizing landscapes of Ohio, USA. Science of the Total Environment 485:461-467. Royer, T. V. and G. W. Minshall. 1997. Temperature patterns in small streams following wildfire. Archiv Fur Hydrobiologie 140:237-242. Rugenski, A. T. and G. W. Minshall. 2014. Climate-moderated responses to wildfire by macroinvertebrates and basal food resources in montane wilderness streams. Ecosphere 5. Russell, R. E., J. A. Royle, V. A. Saab, J. F. Lehmkuhl, W. M. Block, and J. R. Sauer. 2009. Modeling the effects of environmental disturbance on wildlife communities: avian responses to prescribed fire. Ecological Applications 19:1253-1263. Russo, D. and G. Jones. 2003. Use of foraging habitats by bats in a Mediterranean area determined by acoustic surveys: conservation implications. Ecography 26:197- 209. Ryan, S. E., K. A. Dwire, and M. K. Dixon. 2011. Impacts of wildfire on runoff and sediment loads at Little Granite Creek, western Wyoming. Geomorphology 129:113-130. Saab, V. 1999. Importance of spatial scale to habitat use by breeding birds in riparian forests: a hierarchical analysis. Ecological Applications 9:135-151. Saab, V. A. and H. D. W. Powell. 2005. Fire and avian ecology in North America: process influencing pattern. Studies in Avian Biology:1-13. Saab, V. A., R. E. Russell, and J. G. Dudley. 2009. Nest-site selection by cavity-nesting birds in relation to postfire salvage logging. Forest Ecology and Management 257:151-159.

309

Sabo, J. L. and M. E. Power. 2002. Numerical response of lizards to aquatic insects and short-term consequences for terrestrial prey. Ecology 83:3023-3036. Sabo, J. L. and M. E. Power. 2002. River-watershed exchange: effects of riverine subsidies on riparian lizards and their terrestrial prey. Ecology 83:1860-1869. Sabo, J. L., J. C. Finlay, and D. M. Post. 2009. Food chains in freshwaters. Pages 187- 220 in R. S. Ostfeld and W. H. Schlesinger, editors. Year in Ecology and Conservation Biology 2009. Sabo, J. L., J. C. Finlay, T. Kennedy, and D. M. Post. 2010. The role of discharge variation in scaling of drainage area and food chain length in rivers. Science 330:965-967. Sabo, J. L., R. A. Sponseller, M. Dixon, K. Gade, T. Harms, J. Heffernan, A. Jani, G. Katz, C. U. Soykan, J. Watts, and J. Welter. 2005. Riparian zones increase regional species richness by harboring different, not more, species. Ecology 86:56-62. Sala, M., M. Soler, and M. Pradas. 1994. Temporal and spatial variations in runoff and erosion in burnt soils. Pages 1123-1134 the Second International Conference on Forest Fire Research, Coimbra, Portugal. Salvarina, I., E. Yohannes, B. M. Siemers, and K. Koselj. 2013. Advantages of using fecal samples for stable isotope analysis in bats: evidence from a triple isotopic experiment. Rapid Communications in Mass Spectrometry 27:1945-1953. Sanzone, D. M., J. L. Meyer, E. Marti, E. P. Gardiner, J. L. Tank, and N. B. Grimm. 2003. Carbon and nitrogen transfer from a desert stream to riparian predators. Oecologia 134:238-250. Sasaki, T., T. Okayasu, J. Undarmaa, and K. Takeuchi. 2008. Changes in vegetation along a grazing gradient in Mongolian rangelands. Journal of Ecology 96: 145- 154 Schier, G. A. 1973. Origin and development of aspen root suckers. Canadian Journal of Forest Research 3:45-53. Schoener, T. W. 1989. Food webs from the small to the large. Ecology 70:1559-1589. Schumacker, R. E. and R. G. Lomax. 2004. A beginner's guide to structural equation modeling. Lawrence Erlbaum and Associates Inc., Mahwah, NJ. Scott, D. F. 1993. The hydrological effects of fire in South-African mountain catchments. Journal of Hydrology 150:409-432. Scott, D. F. and D. B. Vanwyk. 1990. The effects of wildfire on soil wettability and hydrological behavior of an afforested catchment. Journal of Hydrology 121:239- 256.

310

Seibert, J., J. J. McDonnell, and R. D. Woodsmith. 2010. Effects of wildfire on catchment runoff response: a modelling approach to detect changes in snow-dominated forested catchments. Hydrology Research 41:378-390. Seidman, V. M. and C. J. Zabel. 2001. Bat activity along intermittent streams in Northwestern California. Journal of Mammalogy 82:738-747. Sestrich, C. M., T. E. McMahon, and M. K. Young. 2011. Influence of fire on native and nonnative salmonid populations and habitat in a western Montana basin. Transactions of the American Fisheries Society 140:136-146. Shakesby, R. A. 2011. Post-wildfire soil erosion in the Mediterranean: review and future research directions. Earth-Science Reviews 105:71-100. Shakesby, R. A. and S. H. Doerr. 2006. Wildfire as a hydrological and geomorphological agent. Earth-Science Reviews 74:269-307. Shaw, J. R. 2008. Linkages among watersheds, stream reaches, and riparian vegetation in dryland ephemeral stream networks. Journal of Hydrology 350:68. Shipley, B. 1997. Exploratory path analysis with applications in ecology and evolution. American Naturalist 149:1113-1138. Shipley, B. 2002. Cause and correlation in biology: A user's guide to path analysis, structural equations and casual inference. Smithwick, E. A. H., M. G. Turner, M. C. Mack, and F. S. Chapin. 2005. Postfire soil N cycling in northern conifer forests affected by severe, stand-replacing wildfires. Ecosystems 8:163-181. Speir, S. L., M. M. Chumchal, R. W. Drenner, W. G. Cocke, M. E. Lewis, and H. J. Whitt. 2014. Methyl mercury and stable isotopes of nitrogen reveal that a terrestrial spider has a diet of emergent aquatic insects. Environmental Toxicology and Chemistry 33:2506-2509. Spencer, C. N. and F. R. Hauer. 1991. Phosphorus and nitrogen dynamics in streams during a wildfire. Journal of the North American Benthological Society 10:24-30. Spencer, C. N., K. O. Gabel, and F. R. Hauer. 2003. Wildfire effects on stream food webs and nutrient dynamics in Glacier National Park, USA. Forest Ecology and Management 178:141-153. Stanford, J. A., M. S. Lorang, and F. R. Hauer. 2005. The shifting habitat mosaic of river ecosystems. Verh. Internat. Verein. Limnol. 29:123-136. Stanley, E. H., S. M. Powers, and N. R. Lottig. 2010. The evolving legacy of disturbance in stream ecology: concepts, contributions, and coming challenges. Journal of the North American Benthological Society 29:67-93. Steinmetz, J., S. L. Kohler, and D. A. Soluk. 2003. Birds are overlooked top predators in aquatic food webs. Ecology 84:1324-1328.

311

Stephens, S. L., R. E. Martin, and N. E. Clinton. 2007. Prehistoric fire area and emissions from California's forests, woodlands, shrublands, and grasslands. Forest Ecology and Management 251:205-216. Stephens, S. L., t. Meixner, M. Poth, B. McGurk, and D. Payne. 2004. Prescribed fire, soils, and stream water chemistry in a watershed in the Lake Tahoe Basin, California. International Journal of Wildland Fire 13:27-35. Stewart-Oaten, A., W. W. Murdoch, and K. R. Parker. 1986. Environmental-impact assessment - psuedoreplication in time. Ecology 67:929-940. Stickney, P. F. 1986. First decade plant succession following the sundance forest fire, Northern Idaho. USDA Forest Service. General Technical Report GTR-INT- 197:26 Stone, K. R., D. S. Pilliod, K. A. Dwire, C. C. Rhoades, S. P. Wollrab, and M. K. Young. 2010. Fuel reduction management practices in riparian areas of the western USA. Environmental Management 46:91-100. Suedel, B. C., J. A. Boraczek, R. K. Peddicord, P. A. Clifford, and T. M. Dillon. 1994. Trophic transfer and biomagnification potential of contaminants in aquatic ecosystems. Reviews of Environmental Contamination and Toxicology 136:21- 89. Sullivan, S. M. P. 2012. Geomorphic-ecological relationships highly variable between headwater and network mountain streams of northern Idaho, United States. Journal of the American Water Resources Association 1-12. Sullivan, S. M. P. and A. D. Rodewald. 2012. In a state of flux: the energetic pathways that move contaminant from aquatic to terrestrial environments. Environmental Toxicology and Chemistry 31:1175-1183. Sullivan, S. M. P. and K. T. Vierling. 2012. Exploring the influences of multiscale environmental factors on the American dipper Cinclus mexicanus. Ecography 35. Sullivan, S. M. P. and M. C. Watzin. 2008. Relating stream physical habitat condition and concordance of biotic productivity across multiple taxa. Canadian Journal of Fisheries and Aquatic Sciences 65:2667-2677. Sullivan, S. M. P. and M. C. Watzin. 2009. Stream-floodplain connectivity and fish assemblage diversity in the Champlain Valley, Vermont, U.S.A. Journal of Fish Biology 74:1394-1418. Sullivan, S. M. P., M. C. Watzin, and W. C. Hession. 2004. Understanding stream geomorphic state in relation to ecological integrity: evidence using habitat assessments and macroinvertebrates. Environmental Management 34:669-673. Sullivan, S. M. P., M. C. Watzin, and W. S. Keeton. 2007. A riverscape perspective on habitat associations among riverine bird assemblages in the Lake Champlain Basin, USA. Landscape Ecology 22:1169-1186.

312

Sullivan, S.M.P., Watzin, M.C., 2008. Relating stream physical habitat condition and concordance of biotic productivity across multiple taxa. Canadian Journal of Fisheries and Aquatic Sciences 65, 2667–2677. Swanson, F. J. 1981. Fire and geomorphic processes. U.S. Forest Service General Technical Report WO-26, Washington, DC. Sweeney, B. W., T. L. Bott, J. K. Jackson, L. A. Kaplan, J. D. Newbold, L. J. Standley, W. C. Hession, and R. J. Horwitz. 2004. Riparian deforestation, stream narrowing, and loss of stream ecosystem services. Proceedings of the National Academy of Sciences of the United States of America 101:14132-14137. Swift, S. M., P. A. Racey, and M. I. Avery. 1985. Feeding ecology of Pipestrellus pipistrellus (Chiroptera: Vespertilionidae) during pregnancy and lactation. II. Diet. Journal of Animal Ecology 54:217-225. Tagwireyi, T. and S. M. P. Sullivan. 2015. Distribution and trophic dynamics of riparian tetregnathid spiders in a large river system. Marine and Freshwater Research. Thompson, R. M. and C. R. Townsend. 1999. The effect of seasonal variation on the community structure and food-web attributes of two streams: implications for food-web science. Oikos 87:75-88. Tockner, K., M. S. Lorang, and J. A. Stanford. 2010. River floodplains are model ecosystems to test general hydrogeomorphic and ecological concepts. River Research and Applications 26:76-86. Tonkin, J. D., R. G. Death, and K. J. Collier. 2013. Do productivity and disturbance interact to modulate macroinvertebrate diversity in streams? Hydrobiologia 701:159-172. Townsend, C. R., R. M. Thompson, A. R. McIntosh, C. Kilroy, E. Edwards, and M. R. Scarsbrook. 1998. Disturbance, resource supply, and food-web architecture in streams. Ecology Letters 1:200-209. Townsend, C. R., S. Doledec, R. Norris, K. Peacock, and C. Arbuckle. 2003. The influence of scale and geography on relationships between stream community composition and landscape variables: description and prediction. Freshwater Biology 48:768-785. Van de Water, K. and M. North. 2010. Fire history of coniferous riparian forests in the Sierra Nevada. Forest Ecology and Management 260:384-395. Van de Water, K. and M. North. 2011. Stand structure, fuel loads, and fire behavior in riparian and upland forests, Sierra Nevada Mountains, USA; a comparison of current and reconstructed conditions. Forest Ecology and Management 262:215- 228. Van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z. Fule, M. E. Harmon, A. J. Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen.

313

2009. Widespread increase of tree mortality rates in the western United States. Science 323:521-524. Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Canadian Journal of Fishery and Aquatic Science 37:130-137. Vaughan, I. P., D. G. Noble, and S. J. Ormerod. 2007. Combining surveys of river habitats and river birds to appraise riverine hydromorphology. Freshwater Biology 52:2270-2284. Vaz, P. G., D. R. Warren, P. Pinto, E. C. Merten, C. T. Robinson, and F. C. Rego. 2011. Tree type and forest management effects on the structure of stream wood following wildfires. Forest Ecology and Management 262:561-570. Verkaik, I., M. Rieradevall, S. D. Cooper, J. M. Melack, T. L. Dudley, and N. Prat. 2013. Fire as a disturbance in Mediterranean climate streams. Hydrobiologia 719:353- 382. Verkaik, I., M. Vila-Escale, M. Rieradevall, and N. Prat. 2013. Seasonal drought plays a stronger role than wildfire in shaping macroinvertebrate communities of Mediterranean streams. International Review of Hydrobiology 98:271-283. Verkaik, I., M. Vila-Escale, M. Rieradevall, C. V. Baxter, P. S. Lake, G. F. Minshall, P. Reich, and N. Prat. In press. Stream macroinvertebrate community responses to fire: are they the same in different fire-prone biogeographic regions? Freshwater Science. Vieira, N. K. M., T. R. Barnes, and K. A. Mitchell. 2011. Effects of wildfire and postfire floods on stonefly detrivores of the Pajarito Plateau, New Mexico. Western North American Naturalist 71:257-270. Vieira, N. K., W. H. Clements, L. S. Guevara, and B. F. Jacobs. 2004. Resistance and resilience of stream insect communities to repeated hydrologic disturbances after a wildfire. Freshwater Biology 49:1243-1259. Vierling, K. T. and V. Saab. 2004. Spatial scaling of woodpecker habitat selection in post-fire environments. Ecological Society of America Annual Meeting Abstracts 89:523-523. Vila-Escale, M., T. Vegas-Vilarrubia, and N. Prat. 2007. Release of polycyclic aromatic compounds into a Mediterranean creek (Catalonia, NE Spain) after a forest fire. Water Research 41:2171-2179. Vogl, R. J. 1973. Effects of fire on plants and animals of a Florida wetland. American Midland Naturalist 89:334-347. Voshell, J. R. 2002. A guide to common freshwater invertebrates of North America. The McDonald and Woodward Publishing Company, Blacksburg, Virginia.

314

VTDEC. 2003. Stream geomorphic assessment handbook: rapid stream assessment - phase 2 field protocols. Vermont Agency of Natural Resources, Department of Environmental Conservation, Water Quality Division, Waterbury, VT, USA. Wallace, J. B., S. L. Eggert, J. L. Meyer, and J. R. Webster. 1997. Multiple trophic levels of a forested stream linked to terrestrial litter inputs. Science 277:102-104. Walters, D. M., K. M. Fritz, and R. R. Otter. 2008. The dark side of subsidies; adult stream insects export organic contaminants to riparian predators. Ecological Applications 18:1835-1841. Walters, D. M., M. A. Mills, K. M. Fritz, and D. F. Raikow. 2010. Spider-mediated flux of PCBs from contaminated sediments to terrestrial ecosystems and potential risks to arachnivorous birds. Environmental Science & Technology 44:2849-2856. Wan, S., D. Hui, and Y. Luo. 2001. Fire effects on nitrogen pools and dynamics in terrestrial ecosystems: a meta-analysis. Ecological Applications 11:1349-1365. Ward, J. V., K. Tockner, D. B. Arscott, and C. Claret. 2002. Riverine landscape diversity. Freshwater Biology 47:517-539. Warfe, D. M., T. D. Jardine, N. E. Pettit, S. K. Hamilton, B. J. Pusey, S. E. Bunn, and P. M. Davies. 2013. Productivity, disturbance and ecosystem size have no influence on food chain length in seasonally connected rivers. PLoS ONE 10.1371/journal.pone.0066240. Warren, C. E. and W. J. Liss. 1980. Adaptation to aquatic environments. Pages 15-40 in R. T. Lackey and L. Nielsen, editors. Fisheries management. Blackwell Scientific Publications, Oxford. Westerling, A. L., B. P. Bryant, H. K. Preisler, T. P. Holmes, H. G. Hidalgo, T. Das, and S. R. Shrestha. 2011. Climate change and growth scenarios for California wildfire. Climatic Change 109:445-463. Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and earlier spring increase western US forest wildfire activity. Science 313:940-943. Westerling, A. L., M. G. Turner, E. A. H. Smithwick, W. H. Romme, and M. G. Ryan. 2011. Continued warming could transform Greater Yellowstone fire regimes by mid-21st century. Proceedings of the National Academy of Sciences of the United States of America 108:13165-13170. Whitlock, C. 2004. Forests, fires and climate. Nature 432:28-29. Whitlock, C., S. L. Shafer, and J. Marlon. 2003. The role of climate and vegetation change in shaping past and future fire regimes in the northwestern US and the implications for ecosystem management. Forest Ecology and Management 178:5- 21. Williams, D. D., L. G. Ambrose, and L. N. Browning. 1995. Trophic dynamic of two sympatric species of riparian spider (Araneae, Tetragnathidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie 73:1545-1553. 315

Williams, M. R. and J. M. Melack. 1997. Effects of prescribed burning and drought on the solute chemistry of mixed-conifer forest streams of the Sierra Nevada, California. Biogeochemistry 39:225-253. Wilson, M. F. and K. M. Hocker. 2008. Natural history of nesting American dippers (Cinclus mexicanus) in southeastern Alaska. Northwestern Naturalist 89:97-106. Wilzbach, M. A., B. C. Harvey, J. L. White, and R. J. Nakamoto. 2005. Effects of riparian canopy opening and salmon carcass addition on the abundance and growth of resident salmonids. Canadian Journal of Fisheries and Aquatic Sciences 62:58-67. Wipfli, M. S. 1997. Terrestrial invertebrates as salmonid prey and nitrogen sources in streams: contrasting old-growth and young-growth riparian forests in southeastern Alaska, USA. Canadian Journal of Fisheries and Aquatic Sciences 54:1259-1269. Wisheu, I. C., M. L. Rosenzweig, L. Olsvig-Whittaker, and A. Shmida. 2000. What makes nutrient-poor mediterranean heathlands so rich in plant diversity? Evolutionary Ecology Research 2:935-955. Wolman, M. G. 1954. A method of sampling coarse river-bed material. Transactions of the American Geophysical Union 35:951-956. Wondzell, S. M. and J. G. King. 2003. Postfire erosional processes in the Pacific Northwest and Rocky Mountain regions. Forest Ecology and Management 178:75-87. Yodzis, P. 1984. Energy flow and the vertical structure of real ecosystems. Oecologia 65:86-88.

316

APPENDIX A: Geographic Coordinates of all Study Locations

317

Table 18. Latitude and longitude of (1) 12 paired reaches and reference reach used in

Chapters 2 and 3, (2) 31 study reaches sampled along a gradient of drainage area used in

Chapter 4, and (3) 27 dipper breeding territories sampled in 2012 and 2013 and used in

Chapter 5. All values are expressed as UTM. Map datum: WGS 84.

Study reach longitude latitude

Paired design

Buena Vista 279575.26 4171281.55

Camp 278081.71 4172260.25

Cascade 261966.54 4180701.79

Chilnualna 269621.00 4160857.00

Coyote 261456.96 4180779.58

Crane 255044.77 4180970.48

Frog 254917.50 4207850.22

Grouse 262267.38 4174579.59

Meadow 278229.85 4172296.76

Middle Tuolumne 253776.35 4196357.00

Mono 273526.48 4173090.51

South Tuolumne 257437.93 4188089.55

Tamarack 258627.23 4182335.07

Continued

318

Table 18 continued

Gradient study reaches

Above Wawona 267869.15 4158426.85

Bunnel 284923.20 4180056.41

Cascade 279101.93 4179197.92

Cathedral Beach 270128.00 4178861.00

El Capitan 266199.09 4178167.25

Foresta 253696.35 4173025.35

Gatehouse 259231.42 4174465.35

Generator 261738.00 4178386.00

Gorge 260124.00 4177119.00

Happy Isles 274579.44 4179333.29

Lyle Peak Fork Lower 293028.40 4175253.68

LYV 282108.32 4180211.78

Merced Lake 287973.45 4179451.76

Merced Peak Fork Lower 292882.86 4174317.19

Merced Peak Fork Upper 290578.23 4169783.75

Moraine Dome 280295.86 4178823.61

Moraine Meadows 286175.00 4163862.00

Nevada 277521.46 4178862.83

Pohono 265016.89 4177702.90

Continued

319

Table 18 continued

Raferty 288933.63 4178664.62

Red Peak Confluence 292237.42 4175667.74

Red Peak Fork Lower 291822.32 4175824.40

Red Peak Fork Upper 288081.48 4170198.99

Sugar Pine 273582.00 4180546.00

Superintendents 271538.52 4180470.23

Swamp Lake 283310.00 4158275.00

Triple Divide Peak Fork Upper 293185.80 4170497.25

Warehouse 251525.76 4172877.44

Washburn 290565.00 4177539.00

Wawona Campground 263693.31 4158727.94

Yosemite View 256586.00 4173786.00

American dipper territories

Bridalveil Creek 268238.48 4175496.00

Bridalveil Falls 266610.81 4177803.88

Bunnel Cascade 281460.59 4179451.56

Cascade Falls 261702.54 4180559.94

Cascade Creek 260964.61 4178837.48

Cathedral Creek 280166.00 4201210.00

Chilnualna Creek 267507.72 4159151.46

Continued

320

Table 18 continued

Clark Creek 279292.00 4173161.00

Covered Bridge 265331.71 4157926.22

El Capitan 268113.07 4178464.93

Illilouette Falls 274727.00 4178419.00

Illilouette Creek (a) 274668.43 4175540.10

Illilouette Creek (b) 275267.79 4174326.20

Middle Tuolumne 253529.88 4196231.36

Morrisson Creek 268679.00 4199325.00

Ottoway Creek 282443.57 4169556.97

Piute Creek 273377.00 4203766.00

Red Peak 292237.42 4175667.74

Regulation Creek 283359.00 4201371.00

Rodger's Creek 276009.00 4201885.00

Superintendent's Bridge 271535.74 4180475.86

Tamarack Creek 260471.73 4180092.09

Snow Creek 276594.86 4182099.26

Tenaya Creek 275171.54 4180560.97

Washburn Lake 290458.00 4177714.00

White Wolf 266844.72 4195632.81

Yosemite Falls 270673.59 4183509.59

321

APPENDIX B: Stream Geomorphology and Sediment Data

322

Table 19. Stream cross-section data for the 12 paired reaches and reference reach on Chilnualna Creek used in Chapters 2 and 3.

Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim

Fire. All distance and elevation measures are in reference to the thalweg (i.e., the thalweg is at 0 m elevation and 0 m distance) and

all values are expressed in m.

Flood-prone width Bankfull width Bankfull height Low-bank height Study reach Bed feature Left Right Left Right Left Right Left Right Buena Vista pool 6.2 6.1 6.2 20.3 2.0 2.0 2.1 2.0

2 Buena Vista riffle 8.0 1.6 8.0 18.0 1.5 1.5 3.6 1.5

Buena Vista step 4.7 6.8 4.7 20.3 1.6 1.6 1.2 1.6 Camp pool 3.5 0.6 3.5 0.6 0.6 0.9 0.9 0.9 Camp riffle 2.4 1.4 2.4 2.1 0.5 1.1 0.3 1.1 Camp riffle 1.7 0.7 1.7 1.2 0.6 0.7 0.6 0.7 Cascade pool 5.1 4.1 3.1 4.1 1.5 1.2 3.1 1.4 Cascade riffle 2.6 9.2 2.6 9.2 1.2 1.0 1.5 0.9 Cascade step 10.4 1.6 2.8 1.6 0.6 0.5 2.0 1.1

Continued b

323

Table 19 continued Chilnualna pool 2.9 4.0 2.9 5.2 1.5 1.6 1.3 0.3 Chilnualna riffle 10.5 3.5 10.5 4.0 0.2 1.0 6.8 1.0 Chilnualna riffle 10.3 2.0 10.3 2.4 0.4 0.9 1.5 0.9 Coyote pool 1.4 2.1 1.4 2.1 0.5 0.7 1.1 1.1 Coyote pool 4.0 3.8 4.0 3.8 0.8 0.9 1.6 1.1 Coyote step 2.7 2.4 2.7 2.8 1.1 0.6 2.0 0.6 Crane pool 6.0 0.6 2.2 0.6 0.5 0.8 0.7 1.2 Crane run 1.9 1.0 1.9 2.0 0.9 1.0 0.3 1.0 Crane step 3.9 0.0 2.1 0.5 0.3 0.4 0.5 0.1

3 Frog pool 10.0 5.9 4.4 5.9 0.6 0.4 3.3 1.7

Frog run 9.7 4.7 8.9 8.8 0.8 1.1 2.3 1.1 Frog step 4.4 3.4 2.1 8.2 0.5 1.1 1.0 1.6 Grouse pool 5.0 2.3 5.0 2.7 0.8 0.9 3.2 0.9 Grouse step 3.8 2.2 3.8 2.2 0.7 0.5 2.6 0.5 Grouse step 2.2 3.8 1.6 4.5 0.5 1.2 0.8 1.2 Meadow deep pool 0.6 0.3 0.6 0.5 0.5 0.5 0.5 0.5 Meadow pool 0.5 1.2 0.5 1.2 0.3 0.4 0.2 0.4

Continued b 324

Table 19 continued Meadow riffle 0.8 0.7 0.8 0.8 0.4 0.3 0.6 0.3 Middle Tuolumne riffle 9.0 4.7 9.0 5.7 1.6 1.1 5.0 1.6 Middle Tuolumne riffle 12.9 8.1 12.9 10.6 1.5 1.5 3.9 1.5 Middle Tuolumne run 10.4 3.8 10.4 4.3 1.3 1.1 5.4 1.1 Mono pool 1.9 2.5 1.9 2.5 0.9 0.7 1.5 0.7 Mono riffle 3.9 2.4 3.9 2.6 0.7 0.3 3.2 0.3 Mono step 2.5 5.1 2.5 5.1 0.8 0.8 1.6 0.8 South Tuolumne pool 3.4 7.9 3.4 10.1 2.2 2.2 1.8 2.2 South Tuolumne step 11.7 8.0 11.7 9.0 1.0 0.9 4.1 1.8

5 South Tuolumne step 11.7 5.1 11.7 8.3 1.3 1.0 2.8 1.0

Tamarack pool 2.8 3.2 2.8 3.5 2.7 2.8 1.1 2.8 Tamarack run 3.8 1.0 2.9 1.2 1.5 0.8 2.2 0.8 Tamarack step 5.0 2.8 5.0 4.8 1.8 1.8 2.2 1.8 Following the Rim Fire Cascade step 11.1 10.5 10.3 12.9 1.4 2.6 1.2 2.6 Cascade pool 9.8 11.4 7.6 14.2 2.3 3.9 1.5 3.9 Cascade riffle 6.0 9.1 2.5 11.3 1.3 1.2 1.9 3.3 Continued

325 b Table 19 continued Frog run 12.7 7.1 9.1 9.6 1.1 0.9 3.5 1.3 Frog pool 6.5 2.2 16.3 3.5 1.1 0.9 3.0 1.2 Frog step 19.0 3.3 17.9 4.7 1.1 0.6 4.3 1.0 Middle Tuolumne riffle 9.1 8.2 3.3 8.5 0.7 0.5 2.7 0.9 Middle Tuolumne run 9.8 2.9 6.4 3.8 1.2 2.1 5.8 2.1 Middle Tuolumne riffle 5.7 6.9 5.7 8.7 1.7 1.2 2.5 1.2 South Tuolumne step 13.8 2.6 12.4 3.9 0.5 0.9 8.3 1.1 South Tuolumne pool 5.5 5.7 5.5 7.3 3.1 1.4 3.5 1.5 South Tuolumne step 11.0 4.2 9.9 6.8 0.9 1.0 2.8 1.4

b 326

Table 20. Wolman pebble counts (3 locations per reach) for the 12 paired reaches and reference reach on Chilnualna Creek used in

Chapters 2 and 3. All values are expressed as number belonging to each size class. Both before and after data are presented for the

study reaches utilized in the before-after-control-impact analysis following the Rim Fire.

Sediment clast size (mm)

Study reach Bed Feature 2 2.8 4 5.6 8 11 16 23 32 45 64 90

Buena Vista pool 5 7 3 4 3 4 12 14 14 5 7 10 Buena Vista riffle 1 3 3 7 5 7 8 12 7 13 9 14

3 Buena Vista step 0 1 5 3 5 3 6 5 11 14 9 16

Camp pool 93 2 3 2 0 0 0 0 0 0 0 0 Camp riffle 12 21 29 32 7 1 0 0 0 0 0 0 Camp rifle 32 17 32 16 2 1 0 0 0 0 0 0 Cascade pool 11 7 10 9 8 1 1 8 10 7 6 4 Cascade pool 5 7 6 7 14 10 15 7 6 3 0 3 Cascade step 3 5 6 9 13 2 2 0 3 3 3 5 Chilnualna pool 9 24 21 14 5 4 5 0 7 1 2 0

Continued b 327

Table 20 continued Chilnualna step 0 1 8 15 11 9 8 7 9 3 5 7 Coyote pool 11 7 16 11 7 0 2 3 7 6 11 5 Coyote pool 13 10 20 24 10 6 5 5 2 1 0 0 Coyote step 0 0 0 0 0 0 0 0 0 0 0 0 Crane pool 12 4 11 11 10 24 14 7 6 1 0 0 Crane run 10 17 14 4 12 12 6 12 7 1 1 0 Crane step 0 0 0 0 0 0 0 0 0 0 0 0 Frog pool 8 2 7 10 7 1 2 2 6 6 13 10 Frog riffle 7 5 14 17 4 3 3 0 6 5 4 3

8 Frog step 0 5 12 14 8 11 12 4 4 2 3 8

Grouse pool 9 5 14 18 9 3 3 6 10 2 2 6 Grouse riffle 6 1 4 8 4 3 13 5 1 9 8 9 Grouse step 4 4 11 6 9 2 1 10 8 6 11 5 Meadow pool 17 17 23 19 13 9 2 0 0 0 0 0 Meadow pool 16 28 20 29 6 1 0 0 0 0 0 0 Meadow run 15 11 19 25 15 13 5 3 0 0 0 0 Middle Tuolumne riffle 9 4 11 8 4 2 7 7 8 19 3 4

Continued

328 b

Table 20 continued Middle Tuolumne run 8 6 12 7 8 5 9 6 8 8 8 2 Middle Tuolumne step 11 9 4 5 6 3 9 6 11 3 5 10 Mono pool 11 7 20 26 4 1 1 2 6 4 2 5 Mono riffle 8 5 8 10 2 3 6 8 15 17 11 5 Mono step 6 2 8 4 6 4 5 8 4 10 8 13 South Tuolumne pool 17 19 12 6 5 2 3 3 1 5 8 2 South Tuolumne step 7 7 4 4 6 6 9 5 3 4 5 5 South Tuolumne step 2 2 12 8 12 4 12 8 5 6 4 2 Tamarack pool 4 5 15 20 19 8 9 3 4 0 1 0

9 Tamarack riffle 5 2 9 7 12 8 4 8 9 5 9 15

Tamarack run 23 20 29 20 9 2 2 0 0 0 0 0 Following the Rim Fire Cascade pool 2 6 9 7 9 8 8 7 13 6 9 3 Cascade pool 5 4 7 10 13 9 5 7 11 9 5 3 Cascade step 13 14 11 7 9 11 5 9 10 9 10 8 Frog pool 3 6 5 11 12 9 9 5 6 8 10 4 Frog riffle 1 13 9 12 8 11 9 9 8 10 8 3

Continued b 329

Table 20 continued Middle Tuolumne riffle 6 3 10 7 5 4 7 10 10 5 6 5 Middle Tuolumne run 5 5 8 4 10 4 10 6 13 18 7 4 South Tuolumne pool 16 9 8 5 7 9 7 8 8 5 3 6 South Tuolumne step 10 11 7 11 11 15 14 9 5 4 9 4 South Tuolumne step 9 8 9 11 13 4 7 3 5 11 9 3

Sediment clast size (mm)

3 Study reach

3 Bed Feature 128 180 256-362 362-512 512-1024 1024-2048

Buena Vista pool 2 4 3 2 2 4 Buena Vista riffle 13 14 0 3 3 0 Buena Vista step 11 6 4 0 3 1 Camp pool 0 0 0 0 0 0 Camp riffle 0 0 0 0 0 0 Camp rifle 0 0 0 0 0 0 Cascade pool 3 2 3 8 1 2 Continued

330 b Table 20 continued Cascade pool 0 0 2 6 2 8 Cascade step 14 11 10 2 0 9 Chilnualna pool 4 5 3 3 1 1 Chilnualna step 6 14 5 2 1 0 Coyote Pool 5 6 1 1 0 1 Coyote Pool 4 1 0 5 1 3 Coyote Step 0 0 0 0 0 0 Crane pool 0 0 0 0 0 0 Crane run 2 2 2 0 0 0

1 Crane step 0 0 0 0 0 0

Frog Pool 5 5 4 4 3 6 Frog Riffle 6 11 4 2 4 2 Frog Step 5 3 3 1 2 3 Grouse pool 3 3 2 2 2 1 Grouse riffle 7 7 4 4 4 3 Grouse step 8 5 3 1 4 2 Meadow pool 0 0 0 0 0 0

Continued b 331

Table 20 continued Meadow pool 0 0 0 0 0 0 Meadow run 0 0 0 0 0 0 Middle Tuolumne riffle 5 2 1 1 4 0 Middle Tuolumne run 1 1 8 1 2 0 Middle Tuolumne step 5 5 1 3 4 2 Mono pool 6 2 3 0 0 0 Mono riffle 3 0 0 0 0 0 Mono step 12 8 1 0 0 1 South Tuolumne pool 2 3 2 1 3 2

2 South Tuolumne step 13 7 9 2 3 1

South Tuolumne step 6 4 0 2 2 9 Tamarack pool 2 1 0 0 0 0 Tamarack riffle 2 4 0 1 0 0 Tamarack run 0 0 0 0 0 0 Following the Rim Fire Cascade pool 3 6 3 1 4 1 Cascade pool 2 5 4 5 0 1

Continued b 332

Table 20 continued Cascade step 7 3 2 1 1 1 Frog pool 6 5 4 2 2 0 Frog riffle 0 6 4 3 2 0 Middle Tuolumne riffle 0 5 13 2 2 0 Middle Tuolumne run 4 6 10 1 1 0 South Tuolumne pool 11 5 2 3 0 1 South Tuolumne step 6 9 6 2 2 0 South Tuolumne step 6 4 2 3 5 2

b 333

Table 21. Percent embeddedness for 50 individual sediment clasts at each of 12 paired reaches and reference reach on Chilnualna Creek used in Chapters 2 and 3.

Both before and after data are presented for the study reaches utilized in the before-after-control-impact analysis following the Rim Fire.

Embeddedness (%) [number belonging to each class] Study reach 0-5 6-25 26-50 51-75 76-100 Buena Vista 16 16 10 8 0 Camp 2 8 2 4 34 Cascade 26 12 6 4 1 Chilnualna 11 9 10 12 8 Coyote 20 5 10 9 5 Crane 0 0 5 3 42 Frog 15 15 7 10 3 Grouse 20 14 9 6 1 Meadow 1 5 5 6 33 Middle Tuolumne 16 7 10 8 10 Mono 9 15 7 4 15 South Tuolumne 13 25 18 2 0 Tamarack 10 16 11 9 4 Following the Rim Fire Cascade 25 9 9 9 7 Frog 15 10 11 5 5 Middle Tuolumne 22 10 7 11 1 South Tuolumne 19 16 7 6 2

334

APPENDIX C: Riparian Vegetation Data

335

Table 22. Ocular estimates (octave class) of cover in the herb layer at each of 12 paired reaches and reference reach on Chilnualna

Creek used in Chapter 2. Both before and after data are presented for the study reaches utilized in the before-after-control-impact

analysis following the Rim Fire. Species codes are the first two letters of the genus followed by the first two letters of the species

(e.g., HEMI is Huechera micrantha; see Appendum A for a complete list of species). Ocular estimates are based on octave classes:

1 (trace), 2 (0-1%), 3 (1-2%), 4 (2-5%), 5 (10-25%), 6 (25-50%), 7 (25-50%), 8 (50-75%), 9 (75-95%), 10 (>95%).

Species code

3 [ocular estimate of cover, (octave class)]

6 Study reach Plot # HEMI HELA ERFO FRCA LUPO MEAR ACMI GECA HIAL THFE GACA Buena Vista 1 0 0 0 0 0 0 0 0 0 0 0 Buena Vista 2 0 0 0 0 0 0 0 0 0 0 0 Buena Vista 3 0 0 4 0 0 0 0 0 0 0 0 Buena Vista 4 0 0 0 0 0 0 0 0 0 0 0 Camp 1 5 0 8 4 6 2 3 3 5 0 0 Camp 2 0 6 0 7 4 0 0 4 7 5 0 Camp 3 0 0 3 4 0 0 0 0 0 5 6

Continued b

336

Table 22 continued Camp 4 0 0 7 3 0 0 0 0 0 0 3 Cascade 1 0 0 0 0 0 0 0 0 0 0 0 Cascade 2 7 0 0 0 0 0 0 0 0 0 0 Cascade 3 10 0 0 0 0 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 0 0 0 0 0 Chilnualna 1 0 0 0 0 0 0 0 0 0 0 0 Chilnualna 2 0 0 0 0 0 0 0 0 0 0 0 Chilnualna 3 0 0 0 0 0 0 0 0 5 0 0 Chilnualna 4 5 0 0 0 0 0 0 0 0 0 0

7 Coyote 1 0 4 0 0 0 0 0 0 0 0 0

Coyote 2 0 0 0 0 0 0 0 0 0 0 0 Coyote 3 0 9 0 0 0 0 0 0 0 0 0 Coyote 4 0 6 0 0 0 0 0 0 0 0 0 Crane 1 0 0 0 0 9 0 0 0 0 0 0 Crane 2 0 9 0 0 0 0 0 0 0 0 0 Crane 3 3 0 0 0 3 0 0 0 0 0 0 Crane 4 0 9 0 0 0 0 0 0 0 0 0 Frog 1 0 0 0 0 0 0 0 0 0 0 0 Continued

b 337

Table 22 continued Frog 2 0 0 0 0 0 0 0 0 0 0 0 Frog 3 0 0 0 0 0 0 0 0 0 0 0 Frog 4 0 0 0 0 0 0 0 0 0 0 0 Grouse 1 9 0 0 0 0 0 0 0 0 0 0 Grouse 2 0 0 0 0 0 0 0 0 0 0 0 Grouse 3 0 0 0 0 0 0 0 0 0 0 0 Grouse 4 0 10 0 0 0 0 0 0 0 0 0 Meadow 1 0 0 0 0 0 0 0 0 0 5 0 Meadow 2 0 0 0 0 0 0 5 0 0 0 0

8 Meadow 3 0 0 0 0 0 0 0 4 0 0 0

Meadow 4 0 0 0 9 0 0 3 0 0 0 0 Middle Tuolumne 1 0 0 0 0 5 0 0 0 0 0 0 Middle Tuolumne 2 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 3 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 4 0 7 0 0 7 0 0 0 0 0 0 Mono 1 0 0 0 0 0 0 0 0 0 0 4 Mono 2 0 0 0 0 0 0 0 0 0 0 0 Mono 3 0 0 0 0 0 0 0 0 0 0 0 Continued

338 b

Table 22 continued Mono 4 0 0 0 0 7 0 4 0 0 0 0 South Tuolumne 1 0 0 0 0 9 0 0 0 0 0 0 South Tuolumne 2 0 6 0 0 9 0 0 0 0 0 0 South Tuolumne 3 0 7 0 0 0 0 0 0 0 0 0 South Tuolumne 4 9 0 0 0 0 0 0 0 0 0 0 Tamarack 1 0 0 0 0 6 0 0 0 0 0 0 Tamarack 2 0 0 0 0 0 0 4 0 6 0 0 Tamarack 3 0 0 0 0 0 0 3 0 0 0 0 Tamarack 4 0 0 0 0 0 0 0 0 0 0 0

Following the Rim Fire Cascade 1 10 0 0 0 0 0 0 0 0 0 0 Cascade 2 10 0 0 0 0 0 0 0 0 0 0 Cascade 3 0 0 0 0 0 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 0 0 0 0 0 Frog 1 0 0 0 0 0 0 0 0 0 0 0 Frog 2 0 0 0 0 0 0 0 0 0 0 0 Frog 3 0 0 0 0 0 0 0 0 0 0 0 Frog 4 0 0 0 0 0 0 0 0 0 0 0

Continued b 339

Table 22 continued Middle Tuolumne 1 6 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 2 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 3 0 0 0 0 5 0 0 0 0 0 0 Middle Tuolumne 4 0 0 0 0 0 0 0 0 0 0 0 South Tuolumne 1 0 0 0 0 0 0 0 0 0 0 0 South Tuolumne 2 4 0 0 0 0 0 0 0 0 0 0 South Tuolumne 3 8 0 0 0 0 0 0 0 0 0 0 South Tuolumne 4 0 0 0 0 0 0 0 0 0 0 0

Species code [ocular estimate of cover] Study reach Plot # MIBI EPAN SOCA JUEN EQAR ANMA VICA PTAQ CARO LUBR CAAP Buena Vista 1 0 0 0 0 0 0 0 0 0 0 0 Buena Vista 2 0 0 0 0 0 0 0 0 0 0 0 Buena Vista 3 0 0 7 5 0 6 0 0 0 0 3 Buena Vista 4 0 0 0 0 0 0 0 0 0 0 0 Camp 1 0 0 0 0 0 0 0 0 0 0 0 Camp 2 0 0 0 0 0 0 0 0 0 0 0 Continued

340 b

Table 22 continued Camp 3 2 0 0 0 0 0 0 4 0 0 0 Camp 4 0 0 0 0 0 0 0 0 0 0 0 Cascade 1 0 0 0 0 0 0 0 0 0 0 0 Cascade 2 0 0 5 0 0 0 0 6 0 0 0 Cascade 3 0 0 0 0 0 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 0 0 0 0 0 Chilnualna 1 0 0 0 0 0 0 0 0 0 0 0 Chilnualna 2 0 0 0 0 0 0 0 0 0 0 0 Chilnualna 3 0 0 0 0 0 0 0 0 0 0 0

1 Chilnualna 4 0 0 0 0 0 0 6 0 0 0 0

Coyote 1 0 0 0 0 0 0 0 0 0 0 0 Coyote 2 0 0 0 0 0 0 0 0 0 0 0 Coyote 3 0 0 0 0 0 0 0 0 0 0 0 Coyote 4 0 0 0 0 0 0 0 0 0 0 0 Crane 1 0 0 6 0 0 0 0 0 0 0 0 Crane 2 0 9 0 0 0 0 0 0 0 0 0 Crane 3 0 3 0 0 0 0 0 10 0 0 0 Crane 4 0 0 0 0 0 0 0 0 0 0 0 Continued

341 b

Table 22 continued Frog 1 0 0 0 4 0 0 4 0 0 0 0 Frog 2 0 0 0 5 0 0 4 0 0 0 0 Frog 3 0 9 0 0 0 0 8 0 0 0 0 Frog 4 0 0 0 3 0 6 7 5 0 0 0 Grouse 1 0 0 0 7 0 0 0 0 0 0 0 Grouse 2 0 0 0 0 0 0 0 7 0 0 0 Grouse 3 0 0 0 0 0 0 0 9 0 0 0 Grouse 4 0 0 0 0 0 0 0 8 0 0 0 Meadow 1 0 0 8 9 7 0 0 0 5 0 0

4 Meadow 2 0 0 8 0 0 0 0 0 0 8 0

Meadow 3 0 0 6 0 3 0 0 0 0 7 0 Meadow 4 0 0 0 0 0 0 0 0 0 6 0 Middle Tuolumne 1 0 0 0 8 0 0 5 0 0 0 0 Middle Tuolumne 2 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 3 0 0 0 6 0 0 0 0 0 0 0 Middle Tuolumne 4 0 0 0 0 0 0 0 0 0 0 0 Mono 1 0 0 6 9 0 0 0 0 0 0 0 Mono 2 0 0 5 0 6 0 0 0 0 0 0 Continued

b 342

Table 22 continued Mono 3 0 0 8 0 4 0 0 0 0 0 0 Mono 4 0 0 0 0 0 7 4 0 0 0 0 South Tuolumne 1 0 0 0 0 0 0 0 5 0 0 0 South Tuolumne 2 0 0 0 7 0 0 0 0 0 0 0 South Tuolumne 3 0 3 0 0 3 0 0 0 0 0 0 South Tuolumne 4 0 4 0 0 0 0 0 8 0 0 0 Tamarack 1 0 0 3 0 0 0 0 0 0 0 0 Tamarack 2 0 0 0 0 0 0 0 0 0 0 0 Tamarack 3 0 0 6 0 0 0 0 0 0 0 0

4 Tamarack 4 0 8 0 0 0 0 0 0 0 0 0

Following the Rim Fire Cascade 1 0 0 0 0 0 0 0 0 0 0 0 Cascade 2 0 0 0 0 0 0 0 0 0 0 0 Cascade 3 0 0 0 0 0 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 0 0 0 0 0 Frog 1 0 0 0 0 0 0 0 0 0 0 0 Frog 2 0 0 0 0 0 0 0 0 0 0 0 Frog 3 0 0 0 0 0 0 0 0 0 0 0

Continued b 343

Table 22 continued Frog 4 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 1 0 0 0 0 8 0 0 0 0 0 0 Middle Tuolumne 2 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 3 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 4 0 0 0 0 0 0 0 0 0 0 0 South Tuolumne 1 0 0 0 0 0 0 0 0 0 0 0 South Tuolumne 2 0 0 0 0 0 0 6 0 0 0 0 South Tuolumne 3 0 0 0 0 0 0 9 0 0 0 0 South Tuolumne 4 0 0 0 0 0 0 0 0 0 0 0

Continued

b 344

Table 22 continued

Species code [ocular estimate of cover, (octave class)] Study reach Plot # HEBI LIPA PHHE MECI grass DOJE Buena Vista 1 0 0 0 0 0 0 Buena Vista 2 0 0 0 0 0 0 Buena Vista 3 0 0 0 0 0 0 Buena Vista 4 0 0 0 0 0 0 Camp 1 0 0 0 0 3 0 Camp 2 0 0 0 0 0 0 Camp 3 0 0 0 0 6 0 Camp 4 0 0 0 0 7 0 Cascade 1 0 0 0 0 0 0 Cascade 2 0 0 0 0 5 0 Cascade 3 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 Chilnualna 1 0 0 0 0 0 0 Chilnualna 2 0 0 0 0 0 0 Chilnualna 3 0 0 0 0 8 5 Chilnualna 4 0 0 0 0 0 0 Coyote 1 0 0 0 0 0 0 Coyote 2 0 0 0 0 4 0 Coyote 3 0 0 0 0 0 0 Coyote 4 0 0 0 0 0 0 Crane 1 0 0 0 0 9 0 Crane 2 0 0 0 0 0 0 Crane 3 0 0 0 0 4 0 Crane 4 8 0 0 7 5 0 Continued

345

Table 22 continued Frog 1 8 0 0 0 6 0 Frog 2 0 0 0 0 0 0 Frog 3 6 0 0 0 0 0 Frog 4 0 0 0 0 0 0 Grouse 1 0 9 0 0 0 0 Grouse 2 0 8 0 0 0 0 Grouse 3 0 0 0 0 0 0 Grouse 4 0 0 0 0 0 0 Meadow 1 0 0 0 0 8 0 Meadow 2 0 0 0 0 7 0 Meadow 3 0 0 0 0 6 0 Meadow 4 0 0 0 0 3 0 Middle Tuolumne 1 0 0 0 0 9 0 Middle Tuolumne 2 0 0 0 0 0 0 Middle Tuolumne 3 0 0 0 0 4 0 Middle Tuolumne 4 0 0 0 0 0 0 Mono 1 0 0 0 0 0 0 Mono 2 0 0 0 0 9 0 Mono 3 0 0 0 0 0 0 Mono 4 0 0 0 0 4 0 South Tuolumne 1 0 0 0 0 3 0 South Tuolumne 2 0 0 0 0 0 0 South Tuolumne 3 0 0 0 0 0 0 South Tuolumne 4 0 0 0 0 0 0 Tamarack 1 0 0 4 0 9 0 Tamarack 2 0 0 5 0 3 0 Tamarack 3 0 0 0 0 0 0 Continued

346

Table 22 continued Tamarack 4 0 0 0 4 3 0 Following the Rim Fire Cascade 1 0 0 0 0 0 0 Cascade 2 0 0 0 0 0 0 Cascade 3 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 Frog 1 0 0 0 0 6 0 Frog 2 0 0 0 0 0 0 Frog 3 0 0 0 0 0 0 Frog 4 0 0 0 0 0 0 Middle Tuolumne 1 0 0 0 0 0 0 Middle Tuolumne 2 0 0 0 0 5 0 Middle Tuolumne 3 0 0 0 0 8 0 Middle Tuolumne 4 0 0 0 0 6 0 South Tuolumne 1 0 0 0 0 0 0 South Tuolumne 2 0 0 0 0 0 0 South Tuolumne 3 0 5 0 0 0 0 South Tuolumne 4 0 0 0 0 0 0

347

Table 23. Ocular estimates of cover in the shrub layer (octave class) at each of 12 paired reaches and reference reach on Chilnualna

Creek used in Chapter 2. Both before and after data are presented for the study reaches utilized in the before-after-control-impact

analysis following the Rim Fire. Species codes are the first two letters of the genus followed by the first two letters of the species

(e.g., BEOX is Betula oxidentalis; see Appendum A for a complete list of species). Ocular estimates are based on octave classes: 1

(trace), 2 (0-1%), 3 (1-2%), 4 (2-5%), 5 (10-25%), 6 (25-50%), 7 (25-50%), 8 (50-75%), 9 (75-95%), 10 (>95%).

Species code

8 [ocular estimate of cover, (octave class)]

Study reach Plot # RISP SASP COSE LOIN VICA BEOX CASE RHOC ALRH RUPA AMAL Buena Vista 1 9 8 0 0 0 0 0 0 0 0 0 Buena Vista 2 4 8 0 0 0 0 0 0 0 0 0 Buena Vista 3 0 7 6 4 0 0 0 0 0 0 0 Buena Vista 4 0 6 0 0 5 0 0 0 0 0 0 Camp 1 3 0 0 0 0 0 0 0 0 0 0 Camp 2 3 0 0 0 0 0 0 0 0 0 0

Continued

348

Table 23 continued Camp 3 7 0 0 0 0 0 0 0 0 0 0 Camp 4 0 0 0 0 0 0 0 0 0 0 0 Cascade 1 0 2 0 0 0 0 0 7 0 0 0 Cascade 2 0 0 5 0 0 0 6 9 0 0 0 Cascade 3 0 4 7 0 0 0 0 3 0 0 0 Cascade 4 0 6 7 0 0 0 0 5 0 0 0 Chilnualna 1 0 7 0 0 0 0 0 8 0 5 0 Chilnualna 2 0 0 0 0 0 0 0 0 0 0 0 Chilnualna 3 0 4 6 0 0 0 0 9 0 0 0

9 Chilnualna 4 0 0 0 0 0 0 0 0 0 0 0

Coyote 1 0 0 0 0 0 0 4 8 0 5 5 Coyote 2 0 0 0 0 0 0 6 7 0 0 0 Coyote 3 0 0 9 0 0 0 0 8 0 0 0 Coyote 4 9 5 6 0 0 0 0 0 0 0 0 Crane 1 0 0 6 0 0 0 0 0 0 0 0 Crane 2 0 0 0 0 0 0 0 0 0 0 0 Crane 3 0 0 0 0 0 0 0 0 0 0 0 Continued

Table 23 continued b

349

Crane 4 0 0 6 0 0 0 0 2 0 0 0 Frog 1 4 7 7 0 0 0 0 6 0 0 0 Frog 2 0 9 7 0 0 0 0 0 0 0 0 Frog 3 4 8 6 0 0 0 0 0 0 4 0 Frog 4 4 7 0 0 0 0 0 3 0 0 0 Grouse 1 7 0 0 0 0 0 0 0 0 0 0 Grouse 2 0 0 0 0 0 0 0 5 0 9 0 Grouse 3 6 0 8 0 0 0 0 0 7 8 0 Grouse 4 0 0 0 0 0 0 0 0 0 8 0

0 Meadow 1 0 0 0 0 0 0 0 0 0 0 0

Meadow 2 0 0 0 0 0 0 0 0 0 0 0 Meadow 3 0 5 0 0 0 0 0 0 0 0 0 Meadow 4 0 0 0 0 0 0 0 0 0 0 0 Middle Tuolumne 1 6 4 5 0 0 0 0 0 0 0 0 Middle Tuolumne 2 0 6 7 0 0 0 0 0 0 0 0 Middle Tuolumne 3 0 7 0 0 0 0 0 9 0 0 0 Middle Tuolumne 4 0 0 7 0 0 0 0 9 0 0 0 Continued

Table 23 continued b

350

Mono 1 0 9 0 0 0 0 0 0 0 0 0 Mono 2 0 8 0 0 0 0 0 0 0 0 0 Mono 3 0 9 0 0 0 0 0 0 0 0 0 Mono 4 4 6 0 5 0 0 0 0 0 0 0 South Tuolumne 1 0 9 6 0 0 0 0 0 0 0 0 South Tuolumne 2 0 5 0 0 0 0 8 0 0 0 0 South Tuolumne 3 0 6 8 0 0 0 0 0 0 0 0 South Tuolumne 4 0 7 6 0 0 0 0 0 0 0 0 Tamarack 1 0 0 0 0 0 0 0 0 0 0 0

1 Tamarack 2 0 0 0 0 0 0 0 0 0 0 0

Tamarack 3 0 0 0 0 0 0 0 0 0 0 0 Tamarack 4 0 0 0 0 0 0 0 0 0 0 0 Following the Rim Fire Cascade 1 0 7 7 0 0 0 0 0 0 5 0 Cascade 2 0 5 6 0 0 0 0 0 0 0 0 Cascade 3 0 0 0 0 0 0 0 8 0 0 0

Continued b

Table 23 continued Cascade 4 0 0 6 0 0 0 8 9 0 0 0

351

Frog 1 3 8 5 0 0 0 0 0 0 0 0 Frog 2 0 7 0 0 0 0 0 0 0 0 0 Frog 3 0 8 6 0 0 0 0 0 0 0 0 Frog 4 0 7 8 0 0 0 0 0 0 0 0 Middle Tuolumne 1 4 6 6 0 0 0 0 0 0 0 0 Middle Tuolumne 2 0 6 8 0 0 0 0 0 0 5 0 Middle Tuolumne 3 0 5 0 0 0 0 0 0 0 0 0 Middle Tuolumne 4 0 5 0 0 0 0 0 0 0 0 0

2 South Tuolumne 1 0 6 0 0 0 0 0 0 0 0 0

South Tuolumne 2 0 6 6 0 0 0 0 0 0 0 0 South Tuolumne 3 0 8 0 0 0 0 0 0 0 0 0 South Tuolumne 4 0 0 0 0 0 0 0 0 4 0 0

Continued

b Table 23 continued

Species code [ocular estimate of cover]

352

Study reach Plot # PREM ROSP RILA SYAL CECU ARMA SPDE Buena Vista 1 0 0 0 0 0 0 0 Buena Vista 2 0 0 0 0 0 0 0 Buena Vista 3 0 0 0 0 0 0 0 Buena Vista 4 0 0 0 0 0 0 0 Camp 1 0 0 0 0 0 0 0 Camp 2 0 0 0 0 0 0 0

3 Camp 3 2 0 5 0 0 0 0

Camp 4 6 0 0 0 0 0 0 Cascade 1 0 0 0 0 0 0 0 Cascade 2 0 0 0 0 0 0 0 Cascade 3 0 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 0 Chilnualna 1 0 0 0 0 0 0 0 Chilnualna 2 0 0 0 0 0 0 0 Chilnualna 3 0 0 0 0 0 0 0

Continued Table 23 continued b Chilnualna 4 0 0 0 0 0 0 5 Coyote 1 0 0 0 0 0 0 0

353

Coyote 2 0 0 0 0 0 0 0 Coyote 3 0 0 0 0 0 0 0 Coyote 4 0 0 0 0 0 0 0 Crane 1 0 0 0 0 0 0 0 Crane 2 6 0 0 0 0 0 0 Crane 3 0 0 0 0 0 0 0 Crane 4 0 0 0 0 0 0 0

4 Frog 1 0 0 0 0 0 0 0

Frog 2 0 0 0 0 0 0 0 Frog 3 6 0 0 0 0 0 0 Frog 4 0 0 0 0 0 0 0 Grouse 1 0 0 0 0 0 0 0 Grouse 2 0 0 0 0 0 0 0 Grouse 3 0 0 0 0 0 0 0 Grouse 4 0 0 0 0 0 0 0 Meadow 1 0 0 0 0 0 0 0

Continued Table 23 continued b Meadow 2 0 0 0 0 0 0 0 Meadow 3 0 0 0 0 0 0 0

354

Meadow 4 0 0 0 0 0 0 0 Middle Tuolumne 1 5 0 0 0 0 0 0 Middle Tuolumne 2 0 0 0 0 0 0 0 Middle Tuolumne 3 0 0 0 3 4 0 0 Middle Tuolumne 4 0 0 0 0 4 4 0 Mono 1 0 0 0 0 0 0 0 Mono 2 6 0 0 0 0 0 0

5 Mono 3 0 0 0 0 0 0 0

Mono 4 0 0 0 0 0 0 0 South Tuolumne 1 0 0 0 0 0 0 0 South Tuolumne 2 0 0 0 0 0 0 0 South Tuolumne 3 0 0 0 0 0 0 0 South Tuolumne 4 0 0 0 0 0 0 0 Tamarack 1 0 0 0 0 0 0 0 Tamarack 2 0 0 0 0 0 0 0 Tamarack 3 0 0 0 0 0 0 0

Continued Table 23 continued b Tamarack 4 0 6 0 0 0 0 0 Following the Rim Fire

355

Cascade 1 0 0 0 0 0 0 0 Cascade 2 0 0 0 0 0 0 0 Cascade 3 0 0 0 0 0 0 0 Cascade 4 0 0 0 0 0 0 0 Frog 1 0 0 0 0 0 0 0 Frog 2 0 0 0 0 0 0 4

3 Frog 3 0 0 0 0 0 0 0

Frog 4 0 0 0 0 0 0 0 Middle Tuolumne 1 0 0 0 0 0 0 0 Middle Tuolumne 2 0 0 0 0 0 0 0 Middle Tuolumne 3 0 0 0 0 0 0 0 Middle Tuolumne 4 0 0 0 0 0 0 0 South Tuolumne 1 0 0 0 0 0 0 0 South Tuolumne 2 0 0 0 0 0 0 0 South Tuolumne 3 0 0 0 0 0 0 0 South Tuolumne 4 0 0 0 0 0 0 0

356

Table 24. Ocular estimates of cover in the tree layer (octave class) at each of 12 paired reaches and reference reach on Chilnualna

Creek used in Chapter 2. Both before and after data are presented for the study reaches utilized in the before-after-control-impact

analysis following the Rim Fire. Species codes are the first two letters of the genus followed by the first two letters of the species

(e.g., ABCO is Abies concolor; see Appendum A for a complete list of species). Ocular estimates are based on octave classes: 1

(trace), 2 (0-1%), 3 (1-2%), 4 (2-5%), 5 (10-25%), 6 (25-50%), 7 (25-50%), 8 (50-75%), 9 (75-95%), 10 (>95%).

Species code [ocular estimate of cover, (octave class)] Study reach Plot # ABCO PICO PIPO POTRE CONU LIDE ALRH QULO PILA

5 Buena Vista 1 0 7 0 0 0 0 0 0 0

Buena Vista 2 0 3 0 0 0 0 0 0 0 Camp 1 0 8 0 0 0 0 0 0 0 Camp 2 0 7 0 0 0 0 0 0 0 Cascade 1 7 0 0 0 6 0 0 0 0 Cascade 2 0 0 0 0 0 0 0 0 0 Chilnualna 1 7 6 0 0 0 0 0 0 5

Continued b

357

Table 24 continued Chilnualna 2 0 0 0 0 0 0 0 0 0 Coyote 1 8 0 0 0 0 0 0 0 4 Coyote 2 8 0 0 0 0 0 0 0 4 Crane 1 7 0 0 0 0 0 0 0 0 Crane 2 0 0 0 0 0 0 0 0 0 Frog 1 0 0 0 0 0 0 0 0 0 Frog 2 0 0 0 0 0 0 0 0 0 Grouse 1 0 8 4 0 7 8 6 2 0 Grouse 2 0 0 6 5 9 9 0 0 0

8 Meadow 1 0 4 0 0 0 0 0 0 0

Meadow 2 0 0 0 0 0 0 0 0 0 Middle Tuolumne 1 0 0 0 0 0 0 0 0 0 Middle Tuolumne 2 4 4 0 0 0 8 0 0 5 Mono 1 8 6 3 0 0 0 0 0 0 Mono 2 6 7 2 0 0 0 0 0 0 South Tuolumne 1 5 9 0 0 0 9 0 0 5 South Tuolumne 2 0 0 0 0 0 0 0 0 0 Continued

b 358

Table 24 continued Tamarack 1 0 6 0 0 0 0 0 0 0 Tamarack 2 0 0 0 0 0 0 0 0 0 Following the Rim Fire Cascade 1 6 0 0 0 6 0 0 0 0 Cascade 2 0 0 0 0 0 0 0 0 0 Frog 1 0 0 0 0 0 0 0 0 0 Frog 2 0 0 0 0 0 0 0 0 0 Middle Tuolumne 1 0 0 6 6 0 0 0 0 0 Middle Tuolumne 2 0 0 0 0 0 0 0 0 0

9 South Tuolumne 1 0 7 0 0 0 7 0 0 0

South Tuolumne 2 0 0 0 0 0 0 0 0 0

359

APPENDIX D: Benthic Macroinvertebrate Data

360

Table 25. Density (number of individuals per m2) of benthic macroinvertebrates collected by Surber sampler (500 μm mesh, 2-9

samples per reach depending on channel width) at each of the 12 paired reaches and reference reach on Chilnualna Creek used in

Chapter 3. Density values are the mean across all samples for each reach. Both before and after data are presented for the study

reaches utilized in the before-after-control-impact analysis following the Rim Fire.

Density (individuals per m2) by taxon

Study reach

3 bellaria

6 r

Diptera Ephemeroptera Trichoptera Plecoptera Oligochaeta Megaloptera Coleoptera Tu Hydracarina Buena Vista 4.44 6.11 22.33 21.44 0.33 0.67 3.00 0.11 0.00 Camp 1.67 12.67 0.33 12.33 0.00 0.33 1.00 1.00 0.67 Cascade 3.71 2.43 8.57 2.00 0.43 0.14 2.86 0.00 1.57 Chilnualna 1.00 5.50 5.00 0.50 0.00 0.00 2.00 0.00 7.00 Coyote 7.71 8.43 6.57 2.43 0.00 1.86 0.43 0.00 2.57 Crane 6.67 9.33 5.00 7.33 1.67 0.67 3.67 0.00 1.00 Frog 1.17 31.17 8.17 2.83 0.00 0.67 5.00 0.00 5.17 Grouse 2.89 11.89 4.00 19.56 0.33 2.22 0.33 0.11 1.11 Meadow 3.67 8.33 4.67 7.33 0.00 0.00 0.00 0.00 0.67

b Continued

361

Table 25 continued Middle Tuolumne 7.60 17.80 14.20 8.40 0.40 0.00 13.40 0.20 4.80 Mono 5.56 5.00 4.67 5.89 0.11 0.00 1.78 0.11 0.44 South Tuolumne 6.33 7.89 4.11 7.33 0.11 0.00 2.56 0.11 1.11 Tamarack 3.11 12.78 3.56 11.67 0.56 0.44 0.67 0.00 3.22 Following the Rim Fire Cascade 4.00 5.50 19.00 6.50 0.00 1.00 4.00 0.50 8.00 Frog 4.00 16.00 12.50 4.50 0.00 0.50 1.50 0.00 2.50 Middle Tuolumne 9.00 34.50 23.00 6.50 0.00 2.00 22.50 0.00 3.50 South Tuolumne 22.00 24.00 11.00 0.50 0.00 1.50 4.50 0.00 2.50

362

APPENDIX E: Chemical Water-Quality Data

363

Table 26. Chemical water-quality data collected using a YSI Multiparameter Meter, Cole-Parmer, Vernon Hills, Illinois, USA at

each of the 12 paired reaches and reference reach on Chilnualna Creek used in Chapters 2 and 3. All chemical-water quality

measures were made in the summer of 2014. NA signifies no data for that sampling date due to logistical constraints during

sampling. * No flow signifies disconntinuous surface flow. Note that these data were not used in the analysis, but provide ancillary

information.

Temperature Conductivity DO pH Study reach Date Time ° C μS cm-2 μS cm-1 % mg L-1

4 Buena Vista 28-Jun-14 14:00 18.43 24 21 80.2 7.52 5.64

Buena Vista 8-Jul-14 10:15 15.13 31 25 74.8 7.52 6.34 Buena Vista 19-Jul-14 15:00 19.27 34 31 69.5 6.42 6.01 Buena Vista 6-Aug-14 13:30 17.33 40 34 68.3 6.55 6.35 Buena Vista 16-Aug-14 12:25 17.54 44 38 63.1 6.03 6.32 Buena Vista 26-Aug-14 11:50 15.51 47 38 62.4 6.22 6.28 Buena Vista 4-Sep-14 14:37 17.30 49 42 57.1 5.43 6.14 Continued

b 364

Table 26 continued Buena Vista 27-Oct -14 12:05 7.28 76 30 71.9 8.70 6.45 Camp 28-Jun-14 15:30 10.50 60 43 78.3 8.73 6.57 Camp 8-Jul-14 9:15 9.06 68 47 81.4 9.37 6.97 Camp 19-Jul-14 14:00 11.55 70 52 80.5 8.77 7.05 Camp 6-Aug-14 12:05 9.62 72 51 83.3 9.49 7.02 Camp 16-Aug-14 13:55 10.61 72 52 83.1 9.25 7.22 Camp 26-Aug-14 10:35 8.01 73 49 80.8 9.57 7.06 Camp 4-Sep-14 12:57 9.88 71 51 76.0 8.61 7.20 Camp 27-Oct-14 10:55 4.10 66 40 81.1 10.61 6.92 Cascade 27-Jun-14 15:53 14.53 11 9 88.6 9.01 5.52

6 Cascade 9-Jul-14 15:31 16.88 14 12 87.8 8.51 6.13

Cascade 20-Jul-14 11:21 15.23 17 14 84.6 8.49 6.37 Cascade 7-Aug-14 10:42 12.77 19 15 80.0 8.48 6.30 Cascade 17-Aug-14 11:11 12.72 21 16 79.0 8.38 6.53 Cascade 27-Aug-14 8:15 11.58 21 16 77.0 8.37 6.08 Cascade 5-Sep-14 17:07 15.03 22 17 77.2 7.73 6.60 Cascade 29-Oct-14 9:50 4.58 24 15 83.0 10.73 6.37 Continued

b 365

Table 26 continued Chilnualna 29-Jun -14 11:30 15.40 13 10 77.0 7.70 4.95 Chilnualna 7-Jul-14 14:00 15.63 15 12 77.0 7.66 5.61 Chilnualna 18-Jul-14 10:50 16.93 18 15 76.3 7.37 5.81 Chilnualna 5-Aug-14 12:30 16.08 20 17 80.2 7.90 6.13 Chilnualna 15-Aug-14 11:30 14.04 22 18 72.3 7.44 6.30 Chilnualna 25-Aug-14 10:45 11.90 24 18 66.2 7.13 6.25 Chilnualna October NA NA NA NA NA NA NA Chilnualna September NA NA NA NA NA NA NA Coyote 27-Jun-14 16:00 14.15 15 12 88.2 9.06 5.57 Coyote 9-Jul-14 15:07 17.58 20 18 87.6 8.36 6.60

6 Coyote 20-Jul-14 11:56 16.17 24 20 83.5 8.22 6.64

Coyote 7-Aug-14 10:31 14.50 28 22 84.0 8.56 6.61 Coyote 27-Aug-14 8:26 12.56 31 23 74.8 7.96 6.52 Coyote 5-Sep-14 16:26 17.81 32 27 78.0 7.42 6.66 Coyote 28-Oct-14 10:05 6.47 30 19 79.9 9.82 6.86 Coyote August NA NA NA NA NA NA NA Crane 30-Jun-14 9:20 10.86 38 28 75.8 8.38 5.73 Continued

b 366

Table 26 continued Crane 8-Jul -14 18:19 12.17 39 29 56.5 6.06 5.84 Crane 20-Jul-14 18:40 10.88 42 31 58.4 6.45 5.67 Crane 7-Aug-14 19:01 13.66 26 20 31.0 12.34 7.65 Crane 27-Aug-14 14:40 12.10 40 30 60.6 6.52 6.13 Crane 5-Sep-14 14:32 12.60 17 13 99.6 10.60 6.36 Crane 29-Oct-14 11:55 6.54 43 28 60.0 7.37 6.48 Crane August NA NA NA NA NA NA NA Frog 26-Jun-14 14:45 15.59 19 15 88.1 8.73 5.61 Frog 6-Jul-14 11:40 16.42 21 18 88.9 8.70 6.27 Frog 17-Jul-14 11:45 17.80 25 22 88.4 8.41 5.93

Frog 4-Aug-14 14:15 16.08 28 24 85.6 8.31 6.43 Frog 14-Aug-14 14:03 18.67 32 28 86.1 8.04 6.70 Frog 24-Aug-14 13:45 18.26 33 29 84.8 7.98 6.59 Frog 3-Sep-14 11:00 13.58 34 27 78.9 8.21 6.44 Frog 26-Oct-14 11:15 6.48 35 22 79.3 9.76 6.60 Grouse 29-Jun-14 16:20 14.64 46 37 87.5 8.89 6.06 Grouse 8-Jul-14 13:15 14.32 50 40 91.5 9.36 6.66 Continued

b 367

Table 26 continued Grouse 19-Jul -14 17:45 16.11 55 46 85.0 8.36 6.71 Grouse 6-Aug-14 16:45 14.53 56 45 88.4 9.01 6.53 Grouse 16-Aug-14 16:48 14.07 60 47 85.5 8.79 6.82 Grouse 26-Aug-14 15:15 13.05 60 47 86.9 9.14 6.56 Grouse 4-Sep-14 17:50 13.04 64 49 77.2 8.18 6.89 Grouse 29-Oct-14 15:40 7.07 60 40 80.3 9.74 7.27 Meadow 28-Jun-14 14:45 16.05 25 21 79.1 7.80 5.76 Meadow 8-Jul-14 9:45 11.88 30 22 71.2 7.69 6.49 Meadow 19-Jul-14 14:30 15.94 32 26 72.5 7.28 6.42 Meadow 6-Aug-14 12:45 13.79 33 26 63.9 6.62 6.46

8 Meadow 16-Aug-14 13:22 15.33 33 27 60.1 6.01 6.26

Meadow 26-Aug-14 11:15 11.24 35 26 51.5 5.53 5.96 Meadow 4-Sep-14 13:48 14.00* 40* 31* 35.9* 3.71* 5.82* Meadow 27-Oct-14 11:28 4.25* 54* 33* 36.2* 4.69* 6.03* Middle Tuolumne 27-Jun-14 12:00 14.89 34 27 93.7 9.45 6.36 Middle Tuolumne 6-Jul-14 16:30 21.12 37 34 99.8 8.88 6.80 Middle Tuolumne 17-Jul-14 16:00 22.89 39 37 95.0 8.17 6.85 Continued

b 368

Table 26 continued Middle Tuolumne 4-Aug -14 10:15 15.58 43 35 81.8 8.15 6.41 Middle Tuolumne 14-Aug-14 9:19 13.86 44 35 76.4 7.89 6.55 Middle Tuolumne 24-Aug-14 9:50 14.27 44 35 75.2 7.70 6.66 Middle Tuolumne 3-Sep-14 17:30 19.83 44 39 94.5 8.62 7.03 Middle Tuolumne 26-Oct-14 15:11 7.19 42 28 85.8 10.36 6.56 Mono 28-Jun-14 16:40 16.78 31 26 74.5 7.23 6.00 Mono 8-Jul-14 8:00 12.70 35 27 72.2 7.65 6.25 Mono 19-Jul-14 12:50 15.98 39 32 73.4 7.25 6.28 Mono 6-Aug-14 10:45 10.80 42 30 71.3 7.89 6.09 Mono 16-Aug-14 15:30 15.07 43 35 68.7 6.91 6.33

9 Mono 26-Aug-14 9:25 9.87 44 31 69.1 7.82 6.11

Mono 4-Sep-14 11:49 11.35 46 34 64.2 7.02 6.19 Mono 27-Oct-14 14:00 4.71 45 27 70.2 9.07 6.58 South Tuolumne 30-Jun-14 NA 17.91 21 18 88.7 8.41 5.43 South Tuolumne 9-Jul-14 12:26 19.67 23 21 81.7 5.97 NA South Tuolumne 20-Jul-14 16:34 18.99 26 23 86.6 8.03 6.06 South Tuolumne 17-Aug-14 14:55 18.64 28 25 85.6 7.99 6.77 Continued

b 369

Table 26 continued South Tuolumne 17-Aug -14 14:55 18.64 28 25 85.6 7.99 6.77 South Tuolumne 27-Aug-14 12:20 16.30 28 23 82.6 8.10 6.54 South Tuolumne 5-Sep-14 9:05 12.83 28 21 75.7 8.00 6.28 South Tuolumne 28-Oct-14 11:00 4.23 28 17 82.1 10.71 6.38 Tamarack 30-Jun-14 13:00 16.25 24 20 88.0 8.63 5.69 Tamarack 9-Jul-14 18:08 17.64 26 22 64.6 6.16 6.31 Tamarack 20-Jul-14 17:49 14.46 27 21 85.2 8.69 6.13 Tamarack 7-Aug-14 13:26 15.45 25 20 85.7 5.55 6.18 Tamarack 17-Aug-14 13:20 14.07 27 21 85.6 8.81 6.90 Tamarack 27-Aug-14 10:17 8.81 28 20 83.9 9.73 6.46

0 Tamarack 5-Sep-14 10:30 8.35 28 19 82.9 9.73 6.72

Tamarack 29-Oct-14 11:00 3.74 19 12 71.5 9.45 6.46

370

APPENDIX F: Stable Isotope and Contaminant Data

371

Table 27. Naturally-abundant stable isotope ratios of carbon and nitrogen of benthic algae, detritus (stream conditioned leaf litter),

spiders of the family Tetragnathidae, and predatory benthic macroinvertebrates of the orders Plecoptera and Megaloptera collected

at each of the 12 paired reaches used in Chapter 3. “--” indicates that no individuals were found or collected from that study reach.

Detritus Benthic Algae Tetragnathidae Plecoptera Megaloptera

Study reach 13C 15N 13C 15N 13C 15N 13C 15N 13C 15N Buena Vista -26.29 -1.25 -32.73 -0.62 -25.60 4.36 -27.09 2.24 -- -- Camp -27.31 -2.49 -27.74 0.45 -24.66 3.89 -28.60 3.68 -30.36 6.50 Cascade -28.56 -4.09 -27.54 -2.36 -24.82 4.34 -25.05 1.55 -- --

Coyote -27.40 -0.99 -24.29 -2.32 -24.82 3.29 -24.44 1.14 -23.91 2.53 Crane -27.20 1.89 -30.43 4.37 -30.88 6.96 -29.91 -0.25 -- -- Frog -26.92 -2.65 -26.90 -0.35 -24.38 4.02 -25.10 1.91 -23.98 2.14 Grouse -27.16 -1.86 -28.69 -0.21 -24.78 4.32 -24.73 4.46 -25.12 3.85 Meadow -25.39 -0.20 -35.77 0.88 -- -- -33.34 3.96 -- -- Middle Tuolumne -26.95 -2.96 -29.63 -0.57 -23.85 4.52 -25.30 2.72 -25.65 3.96 Mono -28.08 -2.12 -39.61 1.19 -26.74 5.48 -32.23 4.22 -- -- South Tuolumne -27.23 -2.38 -26.75 -3.24 -24.92 2.96 -24.49 0.64 -22.86 1.67 Tamarack -26.30 -2.21 -26.57 1.18 -27.55 5.20 -29.72 4.04 -- --

b 372

Study reach Arsenic Cadmium Lead Mercury Selenium Thallium Buena Vista 85 6688 335 386 1101 28 Camp ------Cascade 72 3214 238 324 1065 20 Coyote 27 3841 244 328 913 29 Crane 29 2101 282 213 826 31 Frog 40 2959 166 406 852 22 Grouse 109 3842 206 427 959 30 Meadow ------Middle Tuolumne 87 4777 156 447 676 23 Mono 79 3454 171 604 1049 26 South Tuolumne 371 5115 320 505 1144 39 Tamarack 69 4686 275 347 983 36

373

Table 29. Naturally-abundant stable isotope ratios of carbon and nitrogen of benthic alage, detritus (stream conditioned leaf litter),

spiders of the family Tetragnathidae, benthic macroinvertebrates of the order Plecoptera, as well as other benthic

macroinvertebrates collected at each of the 31 locations along a gradient of drainage area used in Chapter 4. “--” indicates that no

individuals were found or collected from that study reach.

Detritus Periphyton Tetragnathidae Plecoptera Study reach 13C 15N 13C 15N 13C 15N 13C 15N Above Wawona -27.83 -2.38 -20.52 -1.21 -22.38 3.34 -23.44 1.31

7 Above Wawona ------22.25 3.85 -22.54 1.44

4 Bunnel -24.38 -3.91 -24.47 -1.09 -24.05 2.69 -24.13 2.31 Bunnel ------24.14 2.66 -- -- Cascade -27.13 -2.11 -29.40 -0.68 -24.88 3.74 -26.06 2.58 Cathedral Beach -29.64 -1.82 -26.67 0.73 -25.40 3.73 -- -- El Capitan -29.34 -0.18 -26.98 0.45 -25.32 3.84 -28.84 2.51 Foresta -29.74 -2.41 -13.58 -1.40 -15.12 8.68 -18.49 1.14 Foresta -- -- -10.64 7.71 ------Gatehouse -27.92 -1.64 -10.90 -0.73 -18.80 3.83 -20.00 1.30

Continued b

374

Table 29 continued Generator -25.86 -1.83 -10.90 -0.73 -20.50 3.14 -- -- Gorge -27.99 -1.37 -15.70 -0.08 -23.87 3.30 -- -- Happy Isles -28.56 -1.52 -18.26 -2.82 -22.51 3.43 -19.14 0.09 Happy Isles -27.61 -3.28 -18.79 -2.74 ------Lyle Peak Fork Lower -27.98 -3.42 -17.59 -4.47 -23.34 0.61 -22.62 -0.56 LYV -27.16 -3.04 -23.49 -2.02 -23.86 2.76 -22.50 1.96 LYV ------24.08 2.79 -- -- Merced Lake -25.14 0.92 -20.99 -1.98 -24.48 2.31 -24.87 0.52 Merced Peak Fork Lower -26.84 -3.80 -20.08 -3.87 -24.09 2.29 -23.19 -1.82

5 Merced Peak Fork Upper -26.91 -4.16 -15.07 -5.79 -24.05 1.65 -22.24 -1.48

Moraine Dome -27.18 -1.59 -21.99 -1.78 -25.02 3.63 -24.40 1.60 Moraine Dome ------25.15 3.39 -- -- Moraine Meadows -27.26 -3.90 -24.84 -4.01 -24.76 2.42 -- -- Nevada -27.14 -1.75 -30.39 -0.80 -25.18 4.21 -25.03 2.95 Pohono -28.06 -1.71 -24.81 -0.08 -24.58 4.01 -28.89 1.71 Red Peak Confluence -27.07 -4.70 -21.35 -4.26 -24.00 1.66 -- -- Red Peak Fork Lower -27.49 -5.11 -23.17 -3.51 -25.04 2.80 -23.86 -1.16 Red Peak Fork Upper -26.18 -2.68 -15.35 -3.68 -24.00 2.41 -- --

Continued b

375

Table 29 continued Stoneman -29.40 -1.14 -26.18 0.20 -24.66 3.56 -25.47 0.91 Sugar Pine -27.91 -4.51 -21.28 -2.59 -24.26 2.46 -- -- Superintendents -28.00 -1.67 -26.18 0.20 -24.46 1.17 -27.05 1.86 Swamp Lake -25.33 -4.18 -21.49 -3.10 -23.90 1.58 -- -- Triple Divide Peak Fork Upper -26.65 -4.39 -16.50 -4.41 -24.70 2.69 -22.94 -0.38 Triple Divide Peak Fork Upper ------23.00 -0.98 Warehouse -29.24 2.38 -10.71 5.98 -18.18 6.99 -16.84 5.24 Warehouse ------17.90 7.04 -- -- Washburn -27.77 -2.61 -22.86 -2.96 -23.98 3.22 -25.51 1.36

6 Washburn -27.86 -3.93 -23.23 -3.90 -24.54 2.02 -- --

Wawona Campground -27.58 -1.66 -13.23 1.52 -23.25 3.30 -22.61 4.59 Yosemite View -26.38 -1.34 -21.84 0.60 -20.85 3.56 -- --

Continued

376

Table 29 continued

Other Benthic Macroinvetebrates Study reach 13C 15N Taxon Bunnel -14.13 5.14 Ephemeroptera Cascade -26.04 1.04 Ephemeroptera Cathedral Beach -27.90 3.09 Chloroperlidae Cathedral Beach -26.80 2.67 Taenioptayginae El Capitan -28.14 1.89 Ephemeroptera Foresta -14.94 0.40 Ephemeroptera Generator -20.74 0.65 Baetidae Generator -18.08 2.06 Chloroperlidae Happy Isles -18.02 -0.26 Ephemeroptera LYV -22.52 -0.68 Ephemeroptera LYV -24.70 2.38 Megaloptera Merced Lake -24.18 0.57 Ephemeroptera Nevada -26.34 2.33 Ephemeroptera Red Peak Confluence -22.99 -0.79 Acroneurinae Red Peak Fork Lower -24.26 -2.89 Baetidae Red Peak Fork Lower -24.81 -3.12 Heptageniidae Red Peak Fork Lower -22.81 -0.98 Papaperlinae Stoneman -25.62 2.43 Megaloptera Superintendents -27.70 1.65 Ephemeroptera Swamp Lake -22.67 -1.77 Baetidae Swamp Lake -24.10 -2.22 Heptageniidae Triple Divide Peak Fork Upper -25.14 2.72 Megaloptera Warehouse -13.98 4.84 Ephemeroptera Washburn -26.35 -1.45 Ephemerillidae Washburn -25.97 0.52 Pteronarys

377

Table 30. Naturally-abundant stable isotope ratios of carbon and nitrogen of benthic algae, detritus (stream conditioned leaf litter),

and dipper feces collected at each of the 27 dipper breeding territories identified in 2012 and 2013 and used in Chapter 5. Both raw

and corrected (i.e., from the linear relationship between swallow feces and blood samples collected from 11 individuals along the

Scioto River in Columbus Ohio) values are shown for dipper feces. See Appendum B and Chapter 5 for complete details on

application of correction factor.

Detritus Benthic Algae Dipper feces Dipper feces

7 (corrected) (raw)

8 Study reach Year 3C 15N 13C 15N 13C 15N 13C 15N Bridalveil Creek 2012 -28.52 -1.92 -28.73 -0.73 -26.07 10.23 -27.70 1.93 Bridalveil Falls 2012 -27.79 -1.97 -16.46 -5.72 -24.69 8.96 -25.18 -1.44 Bridalveil Falls 2013 -27.79 -1.97 -21.86 -3.08 -24.96 9.47 -25.68 -0.09 Bunnel Cascade 2012 -24.63 -2.66 -24.02 -2.06 -24.07 10.28 -24.04 2.07 Cascade Falls 2012 -26.81 -1.07 -10.68 -2.48 -24.48 9.67 -24.80 0.46 Cascade Falls 2013 -26.81 -1.07 -13.96 -1.23 -24.76 10.23 -25.31 1.94 Cascade Creek 2012 -28.56 -4.09 -27.24 -2.57 -25.06 8.96 -25.85 -1.42

b Continued

378

Table 30 continued Cascade Creek 2013 -28.56 -4.09 -24.01 0.71 -24.53 11.51 -24.88 5.34 Cathedral Creek 2013 -26.99 -3.60 -25.16 -1.21 -23.55 9.88 -23.10 1.01 Chilnualna Creek 2012 -28.16 -2.18 -29.87 -2.14 -24.36 8.87 -24.57 -1.68 Chilnualna Creek 2013 -26.78 -4.02 -24.84 -2.84 -24.30 9.79 -24.47 0.77 Clark Creek 2013 -27.55 -3.95 -25.79 -1.39 -24.72 9.51 -25.24 0.02 Covered Bridge 2012 -27.46 -1.23 -21.16 -3.24 -25.15 8.95 -26.01 -1.46 El Capitan 2012 -28.55 -0.76 -19.68 -0.33 -25.29 10.48 -26.28 2.61 Illilouette Falls 2013 -27.61 -3.28 -18.79 -2.74 -24.13 9.85 -24.16 0.94 Illilouette Creek (a) 2012 -27.72 -2.29 -20.26 -1.96 -24.19 9.04 -24.26 -1.23

9 Illilouette Creek (a) 2013 -27.72 -2.29 -20.26 -1.96 -23.83 9.54 -23.60 0.12

Illilouette Creek (b) 2012 -28.11 -1.91 -15.28 -1.86 -23.36 9.60 -22.74 0.27 Illilouette Creek (b) 2013 -28.11 -1.91 -20.81 -1.94 -22.94 9.78 -21.97 0.75 Middle Tuolumne 2012 -26.46 -2.54 -24.52 -1.00 -26.30 10.36 -28.12 2.29 Middle Tuolumne 2013 -26.26 -1.01 -27.91 0.46 -25.05 9.80 -25.83 0.79 Morrisson Creek 2013 -26.85 -1.96 -28.06 -0.29 -24.59 10.02 -25.00 1.37 Piute Creek 2013 -26.59 -0.53 -35.59 1.21 -24.10 10.86 -24.10 3.62 Red Peak 2013 -27.07 -4.70 -21.35 -4.26 -24.04 9.34 -23.99 -0.43 Register Creek 2013 -25.85 -3.33 -25.16 -1.21 -23.57 10.02 -23.13 1.37

Continued b

379

Table 30 continued Regulation Creek 2013 -24.71 -3.97 -16.63 -3.22 -22.92 9.59 -21.94 0.23 Rodger's Creek 2013 -26.92 -2.99 -16.12 -1.92 -22.93 9.87 -21.95 0.98 Snow Creek 2013 -28.96 4.47 -21.30 -1.32 -24.38 9.89 -24.61 1.04 Superintendent's Bridge 2012 -28.17 -1.47 -28.38 -1.92 -25.22 9.97 -26.14 1.26 Tamarack Creek 2012 -25.92 -1.93 -14.25 -0.86 -24.10 10.37 -24.09 2.32 Tamarack Creek 2013 -25.92 -1.93 -27.42 2.51 -23.93 9.91 -23.79 1.10 Snow Creek 2012 -26.66 -1.32 -16.07 -2.92 -23.90 9.82 -23.74 0.84 Tenaya Creek 2012 -28.96 4.47 -27.80 -0.94 -25.30 10.07 -26.30 1.51 Red Peak 2012 -25.59 -2.84 -20.44 -3.48 -23.58 9.16 -23.14 -0.90

0 Washburn Lake 2013 -27.86 -3.93 -23.23 -3.90 -24.66 9.50 -25.13 0.01

White Wolf 2012 -26.85 -1.96 -21.39 -2.22 -27.91 9.99 -31.06 1.31 White Wolf 2013 -26.85 -1.96 -25.60 -0.54 -25.42 9.94 -26.51 1.18 Yosemite Falls 2012 -27.22 -3.89 -16.79 -7.84 -24.86 8.95 -25.49 -1.46

380

Table 31. Contaminant loading of American dipper feces expressed as μg kg-1 collected at each of the dipper breeding territories

sampled in 2012 used in Chapter 5.

Study reach Arsenic Cadmium Lead Mercury Selenium Thallium Bridalveil Creek 1479 2728 571 87 585 80 Bridalveil Falls 829 730 215 40 278 32 Bunnel Cascade 5403 2564 767 150 1421 106 Cascade Falls 2145 1535 681 75 676 62 Cascade Creek 1156 3230 1183 83 706 152

8 Chilnualna Creek 1197 30604 1595 94 945 106

Covered Bridge 3477 18060 2084 56 687 84 El Capitan 3868 2605 2435 119 600 107 Illilouette Falls 10643 5577 797 168 1159 121 Illilouette Creek (a) 1068 2227 480 173 1028 85 Illilouette Creek (b) 1830 4812 921 66 558 114 Middle Tuolumne 725 3241 437 113 625 85 Superintendent's Bridge 10924 1885 3613 59 497 93

Continued b 381

Table 31 continued Tamarack Creek 789 2863 1359 82 607 135 Snow Creek 2205 1766 541 44 430 86 Tenaya Creek 4889 1457 1277 100 783 104 Red Peak 10827 3082 1582 134 1112 159 White Wolf 730 1739 1218 57 298 138 Yosemite Falls 735 1544 1430 115 913 65

382

APPENDIX G: Rapid Habitat Assessment (RHA) Data

383

Table 32. Rapid Habitat Assessment (RHA) scores for American dipper breeding territories broken out by habitat parameter used in

Chapter 5. For riverbank and riparian area scores for each bank have been combined into one.

Dipper breeding territory

woody debris woody substrate bed cover and scour deposition channel morphology hydrologic characteristics connectivity riverbanks area riparian score total Bridalveil Creek 18 16 16 15 16 16 18 18 133

3 Bridalveil Falls 20 16 16 11 15 20 20 18 136

Bunnel Cascade 20 20 20 15 18 18 20 15 146 Cascade Creek 20 13 14 17 18 20 14 20 136 Cascade Falls 13 16 16 13 11 15 20 10 114 Cathedral Creek 17 20 20 20 16 20 16 20 149 Chilnualna Creek 20 18 20 18 15 20 18 20 149 Clark Creek 20 20 20 20 17 17 18 18 150 Covered Bridge 17 13 16 12 15 20 12 16 121

Continued b

384

Table 32 continued El Capitan 16 10 10 18 17 20 20 12 123 Illilouette Creek (a) 20 20 18 18 15 17 12 20 140 Illilouette Creek (b) 20 16 16 16 17 20 12 20 137 Illilouette Falls 18 20 16 17 17 20 18 16 142 Middle Tuolumne 17 12 12 15 17 17 18 20 128 Morrisson Creek 20 16 20 20 17 20 20 20 153 Piute Creek 11 20 20 16 16 20 14 20 137 Red Peak 20 15 16 17 18 18 16 20 140 Register Creek 18 20 20 20 17 20 20 20 155 Regulation Creek 9 20 20 20 17 20 18 20 144

5 Rodger's Creek 20 20 18 20 17 20 18 18 151

Snow Creek 16 15 16 15 10 10 20 10 112 Superintendent's Bridge 16 20 20 17 17 20 20 16 146 Tamarack Creek 20 15 20 16 17 17 18 20 143 Tenaya Creek 20 13 13 17 18 12 19 16 128 Washburn Lake 20 20 20 16 19 20 20 18 153 White Wolf 20 14 17 18 16 16 14 20 135 Yosemite Falls 19 18 20 17 16 20 18 12 140

b 385

APPENDIX H: Independent Variables Used in Model Selection

386

Table 33. Predictions of tetragnathid spider responses [density, reliance on aquatically-derived energy, trophic position (TP), and mercury (Hg) body loading] along with independent variables used in model selection for each prediction in Chapter 3. (See chapter text for details relative to predictions). EPT indicates percent of the aquatic invertebrate community from the orders

Ephemeroptera, Plecoptera, and Trichoptera. Fire frequency was calculated by the proportion of catchment burned > 2x since 1930. Fire extent was calculated by the proportion of catchment burned with moderate or high severity in the most recent fire. D50 is median sediment size.

Tetragnathidae Prediction Independent Variables Density will increase Overhanging vegetation (%) Wood - large and small (%) Fire frequency (%) Fire extent (%) Benthic invertebrate density (no. m-2) EPT (% of community) Precipitation (cm month-1)

Reliance on aquatically derived energy - no net effect Overhanging vegetation (%) Wood - large and small (%) Fire frequency (%) Fire extent (%) Benthic invertebrate density (no. m-2)

D50 (mm) Width-to-depth ratio Continued

387

Table 33 continued Precipitation (cm month-1)

TP (mean and SD) will decrease Fire frequency (%) Fire extent (%) Benthic invertebrate density (no. m-2) EPT (% of community)

D50 (mm) Catchment size (km2) Precipitation (cm month-1)

[Hg] body loading will decrease Fire frequency (%) Fire extent (%) Benthic invertebrate density (no. m-2) EPT (% of community)

D50 (mm) Catchment size (km2) Precipitation (cm month-1)

388

APPENDIX I: Conceptual Diagram of Environmental Drivers of Food-Chain Length

389

Ecosystem size Disturbance Resource availability drainage size fire severity, history & extent (not measured directly)

Niche and species diversity Secondary productivity (benthic invertebrate community composition) (benthic invertebrate density)

Habitat quality (riparian vegetation & geomorphology)

Realized food chain length (Tetragnathidae trophic position)

Figure 1. A conceptual diagram showing how post-fire food chain length may be influenced by ecosystem size, disturbance, and resource availability. Figure 29. A conceptual diagram (used to inform Chapter 4 hypotheses) showing how post-fire food-chain length may be

influenced by ecosystem size, disturbance, and resource availability.

b 390

APPENDIX J: Supplementary Maps

391

Figure 30. Streams (starting at 1st order) and trails in Yosemite National Park, California,

USA.

392

Figure 31. Fire perimeters between 1984 and 2012 displayed by year burned within

Yosemite National Park, California, USA.

393

Figure 32. Number of times burned between 1930 and 2012 (e.g., fire frequency) in

Yosemite National Park, California, USA.

394

Figure 33. Most recent year burned from 1930 to 2012 displayed by decade except for the

2007-2012, Yosemite National Park, California, USA.

395

Figure 34. Fire perimeters from 1984-2012 shown by year within the Merced River and

South Fork of the Merced River catchments, Yosemite National Park, California, USA.

396

Figure 35. Nested drainages along a gradient of drainage area within the Merced River and South Fork of the Merced River catchments, Yosemite National Park, California,

USA.

397

APPENDIX K: National Park Permit

398

399

400

401

402

403

June 6th, 2014

404