A Mammal’s Take on the Rapture
Hypothesis, Jacob’s Ladder, and Other
Notions of Doom, Gloom, and Uniform
Change in Alpine Ecosystems.
Rob Klinger
USGS - Western Ecological Research Center
Bishop, California
What does a place
like this…
With critters like
this…
With critters like
Have to do with a
this?
place like this…
Some pretty big leaps in climate-related effects
Yow!
Climate Change & Alpine Ecosystems
• Virtually no data for the
Sierra Nevada!
– Animals or plants
The influence of plants on animals…
The influence of animals on plants
Purpose Of Talk
• Reminder to think beyond
thermometers and rain
gauges
• Climate shifts do not mean
ecology stops in its tracks
• Focus on interactions and
feedbacks between climate
and ecological processes
Structure Of Talk
• A little (critical) background
• Conceptual foundation of
work in Sierra Nevada alpine
zone
• Patterns of alpine mammal
habitat associations,
distribution, and density
The “Rapture Hypothesis” Scenario
(…with apologies to REM)
•
•
•
•
•
•
It’s the end of
the world as
we know it…
It’s the end of
the world as
we know it…
It’s the end of
the world as
we know it…
•
The planet heats up…
Snowline rises…
Alpine mammals are trapped…
Unable to adapt physiologically ….
Habitat changes…
Many populations appear doomed to
disappear…
But some may not fall off the
mountain
And I
feel fine!
But How Likely Is This Scenario?
• Varying environmental
conditions
• Different life history
characteristics
• Different habitat association
patterns
• Different physiological traits
• Phenotypic and behavioral
plasticity
• Consistent or variable
responses to climatic
shifts?
Three Assumptions To Meet For Consistent
Responses Among Species
• Species have similar habitat
use patterns and
physiological tolerances
• Species are restricted in
habitat breadth
• Unable to adapt to change
The Issues & Major Questions
1. Climate is changing…
–
Likely unprecedented in
recent times
Moberg et al. 2005
2. …but we are not blind
– Species and ecosystems in the
Sierra have continuously
responded to large climatic
fluctuations
3. So…
–
–
–
How have vegetation
communities responded?
How have animal species
persisted?
How have animals and
vegetation interacted?
Large fluctuations in temperature for millennia
Sierra Nevada/White Mountain Alpine Mammal
Study
• 7-10 year study
• Multi-species study
–
–
–
–
–
Bighorn sheep
Yellow-bellied marmot
American pika
Belding’s ground squirrel
Golden-mantled ground squirrel
• Multi-scale
– Rangewide
– Regional
– Local
• Estimate:
–
–
–
–
Occupancy
Habitat associations
Density
Demographic rates
• Model:
– Resource selection
– Species distributions
• Climate
• Topography
• Vegetation
– Population dynamics
– Persistence
• Compare:
– Among species
– Among mountain ranges
– Temperature gradient
• Climate
– Physiological constraints
– Physiological opportunities
• Topography
– Geographic constraint to
dispersal
– Macrohabitat features
• Vegetation
– Macrohabitat features
– Forage (nutrition)demography relationship
– Alpine meadows!
Direct and Indirect
Effects From
Different Factors
But Wait…There’s More!
• Not just effect of plants on
animals
• Mammals play extremely
important roles as
herbivores & granivores
• So could they decouple a
potentially climate-driven
transition of meadows to
forest patches?
What Are Implications Of Potential Range Changes
Of Mammals On Vegetation States?
• Functional group
changes
– Herbivory  granivory
dominated system
0.8
0.8
0.7
0.7
0.6
0.6
Strength
• Changes in distribution
of interaction strengths
0.9
Strength
– Changes in relative
abundance
T+1
T
0.9
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
0.0
marfla
ochpri
spebel
Species
spelat
0.0
marfla
ochpri
spebel
Species
spelat
What Vegetation Patterns Can We Expect Because Of
The Functional Roles Of The Mammals?
• Relatively few homogenous
states …
• The climate-driven scenario
• …Or many alternative stable
states (high heterogeneity)?
• Mammals “manage their own
habitat”
Deterministic, Climate-Driven Transition…
Dry
Pine
Meadows
Forest
Wet
Pine
Meadows
Forest
(Pinus
(Astragalus,
albicaulis,
Phlox,
Pinus
Festuca)
contorta)
(Pinus
(Carex,
albicaulis,
Calamagrostis)
Pinus contorta)
Pine Forest
(Pinus albicaulis, Pinus contorta)
Pine Forest
Elevation
(Pinus albicaulis, Pinus contorta)
Temperature
…Or Alternative States & Pathways?
Dry Meadows
Wet Meadows
(Astragalus, Phlox, Festuca)
(Carex, Calamagrostis)
T3, 5, 6 and/or 7
•
•
•
•
•
•
Herbivory
Granivory
Competition
Soil
Topography
Productivity
T3, 5, 6 and/or 7
T1, 2, 4 and/or 8
T1, 2, 4 and/or 8
Pine Forest/Woody Understory
Pine Forest/Herbaceous
Understory
(Pinus, Artemisia)
T2, 4, and/or 6
T5 and/or 7
(Pinus, Carex, Calamagrostis)
So How Are We Going To Get At This?
Change In Land Cover Classes
Remote Sensing Data
Climate Data
Transects
Point Counts
Habitat Sampling
(1972 - Present)
Mammal Distribution
Models
Mammal Density, Occupancy,
Habitat Associations
Unadjusted
Evaluate interaction
strength
Exclosures
Seeding
Adjusted by RSF’s
Projected Meadow
Conversion Models
Grazing-Herbivory Experiments
(Pinus contorta & P. albicaulis)
Unadjusted
Adjusted for biotic interactions
And What Do The Mammals Think Of All
This?
Some Preliminary Data (2008-2009)
• Density estimates
– Sierra Nevada and White Mtns.
– Southern, central, and northern
Sierra Nevada
• Occupancy estimates
• Habitat associations
– Among species
– Relative to availability
• Density and landscape-scale
habitat use
Methods
– 24 variable-distance line
transects (776 km)
• Sierra Nevada
– N = 18
– 10 km
– 4 samples per year
• White Mountains
– N=6
– 1.4 – 7.8 km
– 4 samples per year
• Occupancy and patch-scale
habitat use
– 180 variable-distance point
count locations
• 10 per transect in the Sierra
Nevada
• 4 samples per year
Northern
White’s
Central
Southern
Sampling spans ≈ 3° latitude
and ≈ 1300 m (4500 feet) elevation
Some Simple Preliminary Expectations
• If species are responding primarily to temperature gradients:
– Densities lower in White Mtns. than Sierra Nevada
– Densities in southern Sierra < central Sierra < northern Sierra
• If species are responding primarily to habitat characteristics:
– No consistent difference in density between geographic strata
• If strong species life-history differences:
– Significant difference in habitat associations and occupancy among
species
If Habitat Use Is Similar Among Species
0.0
0.2
0.4
Psi
0.6
0.8
1.0
Spebel
4050000
4100000
4150000
4200000
UTM South ---> North
Similar responses among species (the null model)
If Variation In Habitat Quality Is Greater Between
Than Within Regions But Habitat Use Differs
Among Species
0.8
0.6
0.0
0.2
0.4
Psi
0.0
0.2
0.4
Psi
0.6
0.8
1.0
Spebel
1.0
Ochpri
4050000
4100000
4150000
4200000
UTM South ---> North
4050000
4100000
4150000
4200000
UTM South ---> North
Distinct differences among regions and species
If Variation In Habitat Quality Is Greater Within
Than Between Regions
0.0
0.2
0.4
Psi
0.6
0.8
1.0
Spelat
4050000
4100000
4150000
4200000
UTM South ---> North
No distinct differences among regions
Density Estimates Stratified By Geographic Region
(Transect Data)
Marfla
50
Individuals km
-2
40
30
20
10
0
South
Central
North
SNV
Region
•
•
•
No geographic pattern across Sierra Nevada
Density in White Mountains 3x - 5x greater than
Sierra Nevada
High variability!
–
–
–
CV = 32% - 45%
86% - 98% attributable to encounter rates
High samples sizes (> 250 observations per species)
WM
Sources Of Variation For Density Estimates
90
0.9
80
0.8
70
0.7
60
0.6
CV
1.0
Percent
100
0.5
50
40
Source
Cluster
Detection
Encounter
30
20
10
0
Density
Geographic
Stratification
0.4
0.3
Stratification
Density
Geographic
0.2
0.1
0.0
marfla
ochpri
spebel
Species
spelat
Density Estimates Stratified By Relative Abundance
Marfla
50
Individuals km
-2
40
30
20
10
0
High
Medium
Low
SNV
WM
Region
• 11% - 22% reduction in CV
0.4
0.2
0.0
-0.2
-0.4
-0.6
Smoothed Estimate
0.6
Marfla - Point Counts
4050000
4100000
4150000
4200000
UTM (South ---> North)
What Is This Telling Us?
• Abundance is patchy
across geographic
gradients!
• Should be reflected in
occupancy and habitat
use patterns
Occupancy
• 55% of sites unoccupied by
any species
• Yellow-bellied marmot
– 23.3%
• American pika
– 43.3%
• Belding’s ground squirrel
– 40.0%
• Golden-mantled ground
squirrel
– 33.3%
No Consistent Pattern Of Occupancy Across Regions
0.8
0.6
0.0
0.2
0.4
Psi
0.0
0.2
0.4
Psi
0.6
0.8
1.0
Spebel
1.0
Ochpri
4050000
4100000
4150000
4200000
UTM South ---> North
4050000
4100000
4150000
4200000
UTM South ---> North
Probability of Occupancy
Marfla
Ochpri
Very patchy environment for all species
Habitat
associations
among
species
Mixing of plots in
ordination space
1.0
Conifer (%)
Separation of species in
ordination space
“Unoccupied”
Relationship of species with
different environmental gradients
Aspect
Spelat
Slope
-1.0
1.0
Marfla
Ochpri
Barren (%)
Spebel
Elevation
Meadow (%)
SPECIES
ENV. VARIABLES
SAMPLES
-1.0
Canonical Correspondence Analysis
South
Central
North
Habitat selection at landscape scale is relatively modest for all four smaller species
Marfla - Transect data regions
1.5
1.0
Selection ratios (+/- CI )
2.0
1.5
Manly selection ratios
Type I data
Rock
Shrub
Meadow
Conifer
Conifer
Rock
Shrub
Meadow
0.0
0.5
1.0
0.5
0.0
Selection ratios (+/- CI )
2.0
2.5
Spelat - Transect data region
Habitat selection at patch scale is very strong for all four smaller species
Marfla - Transect data regions
2.0
Manly selection ratios
Type I data
Shrub
Rock
Meadow
0.0
Shrub
Conifer
Rock
Meadow
Conifer
0.5
1.0
1.5
Selection ratios (+/- CI )
3
2
1
0
Selection ratios (+/- CI )
4
Spelat - Transect data regions
Proportion of habitat types at landscape scale varies among regions
Central
Meadow
Rock
Shrub
Water
0.7
Proportion
0.5
0.6
0.6
0.5
0.4
Proportion
0.3
0.3
0.2
0.2
0.1
0.1
0.0
0.0
Conifer
0.4
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Proportion
North
0.7
South
Conifer
Meadow
Rock
Shrub
Water
Conifer
Meadow
Rock
Shrub
Northern region has proportionally more rock and shrub and less conifer
Central region has proportionally more meadow
ΔAICc= 30.6
Water
Rocksouth
Rockcent
Rocknorth
Shrubnorth
Meadnorth
1
1
2
3
5
6
2
3
Selection ratios (+/- CI )
4
7
Marfla - Transect Data Among
Shrubcent
Connorth
Meadsouth
Meadcent
Concent
Shrubsouth
Consouth
0
Connorth
Consouth
Meadnorth
Rocknorth
Shrubcent
Shrubsouth
Rocksouth
Shrubnorth
Rockcent
Meadcent
Concent
Meadsouth
0
Selection ratios (+/- CI )
Habitat selection varies among regions
Spelat - Transect Data Among
Simple Preliminary Expectations Revisited
• If species are responding
primarily to
temperature/precipitation
gradients:
– Densities lower in White Mtns.
than Sierra Nevada
•
– Densities in southern Sierra <
central Sierra < northern Sierra
•
•
Simple Preliminary Expectations Revisited
• If species are responding
primarily to habitat
characteristics:
– No consistent difference in
density between geographic
strata
• Consistent with
expectations
Simple Preliminary Expectations Revisited
• If strong species life-history
effects :
– Significant difference in
habitat associations among
species
• Consistent with
expectations
Three Assumptions To Meet For Consistent
Responses Among Species Revisited
• Species have similar habitat
use patterns and
physiological tolerances?
– Different habitat use patterns
– Physiology definitely different
• Species are restricted in
habitat breadth?
– Very selective at patch scale
– Less so at landscape scale
• Unable to adapt to change?
– Shift in habitat selectivity in
different geographic regions
Interpreting Results
• Environment is patchy for
all species
• Areas of high and low
probability of persistence
• Species appear to have
plasticity in habitat use
• Uniform responses to
change in environment
unlikely
• Core-satellite structure for
populations
So are alpine
mammals doomed
to very dramatic
and depressing
scenarios?
Come on
up
everyone
…it’s cool
up here!
Me
first!
Make
way, here
we come!
Arrgghhh…
This is the
Titanic in
reverse!
…Or Will They Be Singing A Different Tune
(…and I am still apologizing to REM)
It’s the end of
the world as we
know it (in
places)…
And I feel
fine (most
of the time)!
It’s the end of the
world as we know
it (but I was gonna
move anyway)…
It’s the end of
the world as
we know it
(but I have a
great tan)…
From Here To…
• The need for a MUCH stronger
conceptual foundation
• Technical focus is outstripping
conceptual basis
• Some organizing hypotheses
–
–
–
–
Source-sink
Core-satellite
Dispersal limitation
Metacommunities
• A hierarchical view of species
distributions
–
–
• Real climate variables!
– Temperature and precipitation
– Snow
• Temporal variation
– Vital rates
– Abundance
– Habitat use
• Spatial and temporal dynamics of
vegetation
– Structure
– Species composition
• Macro vs. micro-habitat
characteristics
– Food quantity and quality
– Thermoregulation
• Plant-animal interactions
– Are they decoupling woody
colonization of meadows?
And We Still Have A
Long Way To Go
• CDFG
– Tom Stephenson, Lacey Green,
Lora Konde
• NPS
– Sarah Stock, Sylvia Haultain,
Harold Werner, Steve Thompson
• UC and Cal State
– Dirk Van Vuren, Valerie Eviner,
John Perrine, Qinghua Quo, Otto
Alvarez, Bridgett Benedict
• USFS
– Connie Millar
• USFWS
– Brian Croft
• USGS
– Chris Soulard, Steven Ostoja,
Steven Lee, Jen Chase, Lindsay
Swinger, Stacy Huskins, Laurel
Triatik, Matt Brooks, Nate
Stephenson, NCCWSC
Acknowledgements