A SENTINEL MONITORING NETWORK FOR DETECTINGTHE HYDROLOGIC EFFECTS OF CLIMATE CHANGE ON

SIERRA NEVADA HEADWATER STREAM ECOSYSTEMS AND BIOLOGICAL INDICATORS

David HerbstSierra Nevada Aquatic Research LaboratoryUniversity of CaliforniaMammoth Lakesherbst@lifesci.ucsb.edu

Project funded 2010-2011 by Management Indicator Species program of the US Forest Service, Region 5 Joseph Furnish, contract administrator

Staff: Bruce MedhurstScott RobertsMichael BoganIan Bell

Outline Overview:

• How does climate change influence mountain stream hydrology?

• Why is this important to bioassessment and conservation?

• Using models forecasting hydroclimatic risks vs habitat resistance features to design a monitoring network for the Sierra Nevada

• Preliminary results, insights to stressors, and applications of data set

Motivations for study:

• Forest Service mandate: advance and share knowledge about water and climate change, and how to protect and restore aquatic habitats (Furniss et al. 2010. Water, Climate Change & Forests; USFS PNW-GTR-812)

• Provide a reference stream baseline of natural conditions to produce biological health standards for Management Indicator Species program in National Forests of the Sierra (across 7 National Forests)

• Evaluate the extent of reference decline or drift that might occur with effects of climate change, and use for calibration of CA-SWAMP biocriteria standards in the Sierra

• Integrate climate change in planning & assessment of forest management practices using BMIs for USFS R5

• Assist “Vital Signs” and “Inventory & Monitoring” programs of National Park Service (in 3 Sierra National Parks)

• Modeling has provided some important insights and testable hypotheses on how climate change may alter the thermal and hydrologic regime of streams, but these are no substitute for real data on how aquatic life are responding

• Stream flow and water temperature records show that regime changes are already underway >urgent to establish and maintain an ecological detection network

• Mountain ecosystems with pronounced elevation gradients are especially vulnerable in shifting rain/snow transition zones, with habitat compression occurring in headwaters, and altered flow timing and warming occurring everywhere, all creating ecological challenges

• Design of a monitoring network for detecting climate change effects on mountain streams requires use of models and landscape features to predict where and how hydroclimatic conditions will shift

Background

Headwater habitat compression:

*Drying from above *Warming from below

*Reduced habitat area in lower late summer flows

Regulatory Application of Study:Biological water quality assessment programs

• Programs depend on reference streams to serve as standards for assessing impaired biological integrity

• But what if reference stream conditions are not stable and change beyond natural levels of variation in location and time? Assessment becomes a moving target.

High quality reference sites have most to lose:

• If reference values degrade and become more variable, this increases the signal-to-noise ratio and loss in capability of reference condition to detect impairment

• Climate change may result in reference drift >degraded condition lowers the biological standard

• Need for re-calibration of bio-objectives / standards

JA

N

FE

B

MA

R

AP

R

MA

Y

JU

N

JU

L

AU

G

SE

P

OC

T

NO

V

DE

C

Str

ea

m D

isc

ha

rge

A Changing Hydrograph:Shift in the mountain snowmelt flow regime

earlier snowmelt

wetter & more erraticwinter flows

rain-on-snow floods

Developing and anticipated changes with climate warming

earlier & prolonged low summer flows

periodic drying ofperennial streams

Biological responses of native benthic invertebrates to

warming and altered flow regime:

•Life history/phenology: >favors shorter generations,

& loss of long-lived taxa

>more generalist, fewer specialist traits

>shift to common, widespread taxa

•Migration: >movement into headwater refugia?

•Abundance: >warming increases growth for some

but others cannot survive

warming >habitat areas

diminished under low summer flows

(lost from food chain)

•Physiological stress: >loss of sensitive species

•Emigration escape: >drift in currents or flight

•Dormancy: >hunker-down (resistant stages)

•Shift from a perennial stream community to an

intermittent type (fewer, seasonal-adapted taxa)

•Local extirpation of species

temp

3rd-order size watershedsof Sierra Nevada

Reference3rd-order watersheds

(local impacts minimal to none)

Reference selection filter using GIS:mininimum roadedness or land use, no reservoirs, all above 1000 m)

Climate forecast filter:VIC-hydrological model prediction of snowpack and stream flow

Ranked list of watershedsby quartiles of lowest

and highest climate risk

Natural Resistance Filters: rank low to high•Northness Aspect (snowmelt timing, temp, vegetation)•Groundwater contributions (geology/springs)•Riparian cover and meadow area (water storage)

Low RiskHigh Resistance

High RiskHigh Resistance

Low RiskLow Resistance

High RiskLow Resistance

3 watersheds each categorywith differing exposuresand expectations for theinfluence of climate change

Designed as a natural experiment

field reconnaissanceof best candidate sites

DESIGN

VIC modelOutput:

UseForecast

Change asRisk Level

Historic Hydrograph (1950-2000)

0

2

4

6

8

10

12

14

0 30 60 90 120 150 180 210 240 270 300 330 360

Day of Year

Runoff (mm/day) Baseflow (mm/day) Snow Water Equivalent (mm/100)

M odeled Future Hydrograph (2041-2060)

0

2

4

6

8

10

12

14

0 30 60 90 120 150 180 210 240 270 300 330 360

D ay of Year

R unoff (mm/day) B aseflow (mm/day) Snow Water Equivalent (mm/100)

SWE surface flow

baseflow

Surface flow lower& depleted earlier

Ground snow cover less& disappears sooner

Winter flows more variableWith extreme flow events

Baseflow lower

Sentinel Monitoring Networkfor Sierra Nevada

Selections based on summedClimate-Risk factors from VIC:•Reduction in April 1 SWE•Change in total AMJ run-off•Change in total AMJ base-flow

upper quartile of change =high risklower quartile of change =low risk

Natural Resistance:upper / lower quartiles forNorth-facing = low vulnerabilitySouth-facing = high vulnerabilityPlus, resistance conferred by deepgroundwater-recharge potential from basalt / andesite geology area(Tague and others 2008)

12 catchments24 sites total(tributary sitenested in eachcatchment)

17 in 7 National Forests7 sites in 3 National Parks

Nested Tributaries

CatchmentReach 3°

20-100 km2

Tributary Reach1° or 2°

Survey Monitoring data collected:•150-m reach length•channel geomorphologyincluding bankfull cross-sections(substrata-depth-current profiles,embeddedness, slopes, bank andriparian cover, riffle-pool ratio, etc)•conductivity, alkalinity, SiO2, pH•large woody debris inventory•cobble periphyton (Chl a, taxa IDs)•CPOM & FPOM resources•macroinvertebrates (RWB & TRC)•adult aquatic insect sweeps•photo-points

12 catchments + tributary in each:24 stream reaches total network

MonitoringProtocols:

SWAMP-based

Instrumentation set up atmonitoring stations:

Temperature probesat tributary reaches

Stage-level pressure transducersand Temperature probes atcatchment reaches (water and air)

40 min recording intervals

GIS Analysis at each:land use, roads, geology,riparian, meadow & forest cover,cool-air pooling algorithm,groundwater recharge

Tyndall

Upper Bubbs

not enough $ to putpressure transducersin tributaries

Elevation and Region

GPS Latitude

36 37 38 39 40 41 42

Ele

va

tio

n (

m)

0

1000

2000

3000

4000

Streams in the southern region are atmid-to-high elevations, with low levelsof conductivity and dissolved silicate(snow-melt dominated)

Streams in the northern region are atlower elevations, with higher levelsof conductivity and silicate(groundwater mineral content)

Northern streams support higher levelsof biological diversity than in the south

Taxa Richness and Region

GPS Latitude

36 37 38 39 40 41 42

Taxa R

ich

ness

20

30

40

50

60

70

Chem istry and Region

GPS Latitude

36 37 38 39 40 41 42

Co

nd

uc

tiv

ity (

uS

)

0

50

100

150

200

250

C hem istry and R eg ion

G PS Latitude

36 37 38 39 40 41 42

SiO

2

0

10

20

30

40

w/oM&Ms

Climate Change

Distance (Objective Function)

Information Remaining (%)

3.9E-03

100

8.1E-02

75

1.6E-01

50

2.3E-01

25

3.1E-01

0

Butte

Willow

McCloud

EF Moosehead

EF Nelson

Cat

Robinson

Grassy Swale

MFCosumnes

Sagehen1

Sagehen2

Nelson

Warner

Cathedral

Upper Cathedral

Deer

Pitman

Snow Corral

Crown

Tyndall

Upper Bubbs

Forester

Upper Tyndall

SF Tamarack

North ofYosemite

Yosemiteand South

Intermittent channel = shortest upstream length, low SiO2 snowmelt

Community Similarity Groupings

Nearly300 taxaidentifiedto dateIncl.M&Ms

Closer look at Intermittent flow: stress of periodic summer drying>perennial upstream length used as indicator of dependable flow

Short headwater streams most susceptible, having least taxa richness.But what protects some headwaters and not others?>Groundwater inflows sustain baseflow (higher SiO2) and resist drying>low SiO2 snowmelt risk drying but support more richness with increased channel length (=perennial flow)

0

10

20

30

40

50

60

70

0.1 1 10 100 1000

Upstream Channel Length (km)

(lower values risk drying)

Ta

xa

Ric

hn

ess

0

10

20

30

40

50

60

70

0.1 1 10 100 1000

Upstream Channel Length (km)

(lower values risk drying)

Ta

xa

Ric

hn

es

s

Silicate < 15 mg/L

Silicate > 15 mg/L

EP

T R

ich

ness

0

10

20

30

40

50

Risk = High High Low LowVulnerability = High Low High Low

EPT Richness and Treatment Category

+ 1 SD

Statistical equivalence of treatment groups:Group with most risk and vulnerability alsohas the most scope for response.All groups exceed the Sierra reference levelfor maximum EPT = 23 taxa [eastern Sierra IBI]

Biodiversity present in treatment groups have similar initial richness levelsfor 2010, a near-average water year

=Baseline forfurther comparisons

Trait Character States:•79% of these taxa are cold-adapted*•89% are either semi- or uni-voltine (have ≥1 yr life cycles)•67% prefer riffle habitat (high flows)

*except intermittent stream just 50%

Flow Regime Types Observed*(habitat ecological templates, after Poff and others)

Are there associated BMI community types?

• 1. Stable winter flows and temperatures during ice cover (though R on S may occur), rapid spring snow-melt and summer recession, prolonged cool temps (<10ºC)

• 2. Winter rain and snow, instable ice-snow cover, rising flows through winter and spring, warm summer temperatures (≥15ºC)

• 3. Stable groundwaters sustain high flows and cooler more constant temperatures (≤10ºC)

• 4. Spatial intermittent flows, losing reaches, warm, variable

Upper Bubbs Creek 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

De

pth

(m

)

0.0

0.5

1.0

1.5

2.0

Upper Bubbs Creek 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

Te

mp

era

ture

(C

)

-5

0

5

10

15

20 Deer Creek 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

Tem

pera

ture

(C

)

-5

0

5

10

15

20

Deer Creek 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

Dep

th (

m)

0.0

0.5

1.0

1.5

2.0McCloud River 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

Dep

th (

m)

0.0

0.5

1.0

1.5

2.0

McCloud River 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

Tem

pera

ture

(C

)

-5

0

5

10

15

20

Cathedral Creek 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

Dep

th (

m)

0.0

0.5

1.0

1.5

2.0

Cathedral Creek 2010-2011

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug

Tem

pera

ture

(C

)

-5

0

5

10

15

20

*1. Snow 2. Rain+Snow 3. Groundwater 4. Intermittent-Flashy

6.0

6.5

7.0

7.5

8.0

8.5

100 1,000 10,000 100,000

Density (ind/m2)

pH

2010 data

6.0

6.5

7.0

7.5

8.0

8.5

0 10 20 30 40 50

EPT Diversity

pH

2010 data

2011 prelim data shows no loss of diversity/abundance: resilient so far

2011: high and prolonged spring runoffwater chemistry change: lower pH (-0.75 mean)Wilcoxon signed-rank paired comparison 2010 to 2011p<0.0001 (22 of 24 streams), decreasing from an average of 7.22 to 6.47

pH decrease with high runoff dilution of inflows, most severe at streams with lower pH and less acid neutralizing capacity

Biological Consequences? ...depending on persistence of lower pH

What’s next: using the data obtained and maintaining the network into the future

• Sustain funding - possibly through interagency cost-sharing?• Contribute results to California Climate Change Portal,

and integrate into assessment reports of US-GCRP• Apply flow and temperature recordings for the past year to

validate and calibrate ungaged flow models, and use to back-cast past flow histories (use by USGS, DWR?)

• Further analysis of 2011 data to evaluate reference stability and biological indicator responses to record snowfall, high runoff, reduced pH, and delayed spring onset

• Do communities correspond to hydrographic regimes?

Invite Collaboration• High elevation hydro- and thermo-graphs for model

development >>rare data from headwater streams to share• Stable isotope analysis of heavy water (18O & 2H) at each site

to determine groundwater contribution (mixing models)

Conservation Applications

• Although there are many endemic and montane-adapted native species of aquatic inverts in the Sierra Nevada, biogeography and habitat requirements are poorly known, so surveys supply a basic biodiversity inventory

• Improved understanding of natural flow and temperature regimes, and microclimate of headwater streams

• Identify habitats & taxa changing most, and how these might be protected from climatic effects on hydrologic and thermal regimes > refugia & aquatic diversity management areas

• Extend GIS analysis of environmental resistance factors to assess habitat sensitivity to climate risk

• Use ecological trait analysis to assess biotic vulnerability• Develop management framework to prioritize stream types

for building resilience and protection of most vulnerable watersheds (riparian & meadow restoration, protect groundwater infiltration paths, reduce soil loss/debris flows by managing grazing, logging & road disturbance)

• USFS-NPS adaptive planning in climate change stewardship

conclusions• Network is up and running and the biological indicators

provide a strong foundation for detecting change (biodiversity & trait sensitivities to hydro-climate change)

• North – South stream groups show distinct differences in snowmelt vs groundwater influence on hydrology (and related water chemistry), and biological communities

• Biological richness of northern streams is ecological “insurance”, but this also means more to lose in a region with the most severe climate risk predicted

• Though having less biodiversity, southern streams harbor some vulnerable taxa with restricted distributions

• Intermittent drying poses a clear risk to sustaining biodiversity, esp. in snowmelt-dominated streams, but groundwater systems appear to be more buffered (confirming a predicted climate risk-resistance)

• Lower pH an unexpected change with uncertain effects under extreme flow conditions, but so far communities of BMIs appear stable under this stress

Questions?