Brook trout popula-on dynamics: Integrated modeling across scales ...
Ben Letcher, Yoichiro Kanno, Ron Bassar, Ana Rosner, Paul Schueller, Kyle O’Neil, Krzysztof Sakrejda, MaQ O'Donnell, Todd Dubreuil Conte Anadromous Fish Research Center, U.S. Geological Survey, Turners Falls, MA, USA Keith H. Nislow, Jason Coombs Northern Research Sta+on, USDA Forest Service, Amherst, MA, USA Andrew Whiteley Department of Natural Resources Conserva+on UMass, Amherst, MA, USA
Steve Hurley
Brook trout popula+on dynamics: Integrated modeling across scales and data types
Overview è
è
Goal: understand popula+on dynamics and provide broad spa+al scale forecasts in response to environmental change Problem: specificity/generality tradeoff Can’t do detailed, mechanis+c studies everywhere ¤ Lots of good survey data ¤
è
Approach/solu+on: combined approach
How does/will environmenta l change affect stream salmonids?
Data types è
PIT tag ¤
Single-‐site demographic models n
è
Abundance ¤
Mul+ple-‐site demographic models n
è
Seasonal sensi+vity of lambda (popula+on growth)
Sensi+vity + basin characteris+cs
Presence/absence ¤
Occupancy models n
Effects of long term means + basin characteris+cs
Data types è
PIT tag ¤
Single-‐site demographic models n n n
è
Abundance ¤
Mul+ple-‐site demographic models n n n
è
Sensi+vity of lambda Strong seasonal results Pathways of sensi+vity
Sensi+vity of lambda Seasonal results with enough data Basin characteris+cs effects
Presence/absence ¤
Occupancy models n n
Basin characteris+cs effects Long-‐term mean effects
What can we es+mate? Data type
Model
Endpoint
Basin characteris 4c effects?
Yearly environment al effects
Seasonal environment al effects
Pathways of seasonal environment al effects
PIT tag
Single-‐site demographic
Popula+on growth
No
Yes
Yes
Yes
Abundance
Mul+ple-‐site demographic
Popula+on growth
Yes
Yes
Yes, but need lots of data
No
P(occupancy)
Yes
No
No
No
Pres/abs
Occupancy
Data types è
PIT tag ¤ ¤ ¤
è
Abundance ¤
è
Single-‐site demographic model Body growth, survival, movement, reproduc+on Integral projec+on model
Abundance models
Presence/absence ¤
Occupancy models
West Brook
Isolated
Data types è
Presence/absence ¤ Occupancy models
è
Abundance ¤ Abundance models
è
PIT tag ¤ Mechanis+c models
Autumn
Lambda sensi+vi+es
Spring ↔ Winter ↔ Autumn ↓ Summer ↓
Summer↑ Autumn ↑ Spring ↔ Winter↓
Lambda response surfaces
Forecast
Data types è
PIT tag ¤
è
Single-‐site demographic model
> age-‐0+
Abundance ¤
Abundance models n n n
¤ ¤
è
Age-‐0+
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Yearly data, many sites
All
Es+mated abundances è
PIT tag ¤
è
Single-‐site demographic model
Abundance ¤
Abundance models n n n
¤ ¤
è
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Forecast è
PIT tag ¤
è
Single-‐site demographic model
Abundance ¤
Abundance models n n n
¤ ¤
è
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Forecasts è
Presence/absence
¤ Occupancy models
è
Abundance
¤ Abundance models ¤ Simple popula+on
projec+on -‐ state space
è
PIT tag
¤ Mechanis+c models
↑ ↓ ↔
Extreme events forecast è
PIT tag ¤
è
Single-‐site demographic model
Abundance ¤
Abundance models n n n
¤ ¤
è
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Data types è
PIT tag ¤ Single-‐site
demographic model
è
Abundance
¤ Abundance models
è
Presence/absence ¤ Occupancy models
Single or mul+ple year data, many sites
Model es+mates è
PIT tag ¤
è
Precip
% forest
Abundance ¤
è
Single-‐site demographic model
Abundance models
Presence/absence ¤
Occupancy models n n n n n n
Annual precipitation Minimum temperature Soil drainage class Drainage area Forest cover Stream slope
Air T
Probability of Occupancy for Current Conditions
Brook Trout Probability of Occupancy < 10% 11% -‐ 20% 21% -‐ 30% 31% -‐ 40% 41% -‐ 50% 51% -‐ 60% 61% -‐ 70% 71% -‐ 80%
Model drivers Drainage area Forest cover Stream slope
Annual precipitation Minimum temperature Soil drainage class
81% -‐ 90% > 90%
Probability of Occupancy for Current Conditions Probability of Occupancy 2 C increase
Brook Trout Probability of Occupancy < 10% 11% -‐ 20% 21% -‐ 30% 31% -‐ 40% 41% -‐ 50% 51% -‐ 60% 61% -‐ 70% 71% -‐ 80%
Model drivers Drainage area Forest cover Stream slope
Annual precipitation Minimum temperature Soil drainage class
81% -‐ 90% > 90%
Probability of Occupancy 4 C increase
Probability of Occupancy for Current Conditions
Resilience of occupancy to temperature increase
Brook Trout Probability of Occupancy < 10% 11% -‐ 20% 21% -‐ 30% 31% -‐ 40% 41% -‐ 50% 51% -‐ 60% 61% -‐ 70% 71% -‐ 80%
Model drivers Drainage area Forest cover Stream slope
Annual precipitation Minimum temperature Soil drainage class
81% -‐ 90% > 90%
Brook Trout Resilience Increase tolerated (°C) 0° 0.1° -‐ 0.5° 0.6° -‐ 1° 1.1° -‐ 1.5° 1.6° -‐ 2° 2.1° -‐ 2.5° 2.6° -‐ 3° 3.1° -‐ 8 .5°
Currently below threshold
Bringing it together Variable
Flow
Temperature
Season
Model Single-‐site demographic
Mul+ple-‐site demographic
Occupancy
Fall
↑ **
↑ ***
Winter
↓ **
↓ **
Spring
↔
↔
Precip ↑
Summer
↑ ***
NA
Fall
↓ **
NA
Winter
↔
NA
Spring
↔
↔
Summer
↓ ***
NA
Temperature ↓
Summary è
Congruent environmental effects on popula+on growth across scales
Increases confidence in generality of results ¤ Nega+ve effects of temperature ¤ Posi+ve effects of flow in fall and summer, nega+ve effects in winter ¤
è
Many brook trout popula+ons at risk in future Flow and temperature ¤ Extreme events ¤
è
Can iden+fy resilient popula+ons
Steve Hurley
Web app è
Map viewer ¤ Standard layers ¤ Data ¤ Model results ¤ Select a basin for scenario tester
è
Scenario tester ¤ Climate -‐> Landuse -‐> Environment -‐> Popula+on
response ¤ Evaluate management ac+ons under alternate futures hQp://felek.cns.umass.edu:8080/geoserver/www/data.html
Data types Sensi+vity of annual survival
Steve Hurley
Brook trout popula+on dynamics: Integrated modeling across scales and data types
Overview è
è
Goal: understand popula+on dynamics and provide broad spa+al scale forecasts in response to environmental change Problem: specificity/generality tradeoff Can’t do detailed, mechanis+c studies everywhere ¤ Lots of good survey data ¤
è
Approach/solu+on: combined approach
How does/will environmenta l change affect stream salmonids?
Data types è
PIT tag ¤
Single-‐site demographic models n
è
Abundance ¤
Mul+ple-‐site demographic models n
è
Seasonal sensi+vity of lambda (popula+on growth)
Sensi+vity + basin characteris+cs
Presence/absence ¤
Occupancy models n
Effects of long term means + basin characteris+cs
Data types è
PIT tag ¤
Single-‐site demographic models n n n
è
Abundance ¤
Mul+ple-‐site demographic models n n n
è
Sensi+vity of lambda Strong seasonal results Pathways of sensi+vity
Sensi+vity of lambda Seasonal results with enough data Basin characteris+cs effects
Presence/absence ¤
Occupancy models n n
Basin characteris+cs effects Long-‐term mean effects
What can we es+mate? Data type
Model
Endpoint
Basin characteris 4c effects?
Yearly environment al effects
Seasonal environment al effects
Pathways of seasonal environment al effects
PIT tag
Single-‐site demographic
Popula+on growth
No
Yes
Yes
Yes
Abundance
Mul+ple-‐site demographic
Popula+on growth
Yes
Yes
Yes, but need lots of data
No
P(occupancy)
Yes
No
No
No
Pres/abs
Occupancy
Data types è
PIT tag ¤ ¤ ¤
è
Abundance ¤
è
Single-‐site demographic model Body growth, survival, movement, reproduc+on Integral projec+on model
Abundance models
Presence/absence ¤
Occupancy models
West Brook
Isolated
Data types è
Presence/absence ¤ Occupancy models
è
Abundance ¤ Abundance models
è
PIT tag ¤ Mechanis+c models
Autumn
Lambda sensi+vi+es
Spring ↔ Winter ↔ Autumn ↓ Summer ↓
Summer↑ Autumn ↑ Spring ↔ Winter↓
Lambda response surfaces
Forecast
Data types è
PIT tag ¤
è
Single-‐site demographic model
> age-‐0+
Abundance ¤
Abundance models n n n
¤ ¤
è
Age-‐0+
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Yearly data, many sites
All
Es+mated abundances è
PIT tag ¤
è
Single-‐site demographic model
Abundance ¤
Abundance models n n n
¤ ¤
è
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Forecast è
PIT tag ¤
è
Single-‐site demographic model
Abundance ¤
Abundance models n n n
¤ ¤
è
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Forecasts è
Presence/absence
¤ Occupancy models
è
Abundance
¤ Abundance models ¤ Simple popula+on
projec+on -‐ state space
è
PIT tag
¤ Mechanis+c models
↑ ↓ ↔
Extreme events forecast è
PIT tag ¤
è
Single-‐site demographic model
Abundance ¤
Abundance models n n n
¤ ¤
è
Autumn, Winter, Spring Flow Spring Temperature Eleva+on
State space Popula+on projec+on
Presence/absence ¤
Occupancy models
Data types è
PIT tag ¤ Single-‐site
demographic model
è
Abundance
¤ Abundance models
è
Presence/absence ¤ Occupancy models
Single or mul+ple year data, many sites
Model es+mates è
PIT tag ¤
è
Precip
% forest
Abundance ¤
è
Single-‐site demographic model
Abundance models
Presence/absence ¤
Occupancy models n n n n n n
Annual precipitation Minimum temperature Soil drainage class Drainage area Forest cover Stream slope
Air T
Probability of Occupancy for Current Conditions
Brook Trout Probability of Occupancy < 10% 11% -‐ 20% 21% -‐ 30% 31% -‐ 40% 41% -‐ 50% 51% -‐ 60% 61% -‐ 70% 71% -‐ 80%
Model drivers Drainage area Forest cover Stream slope
Annual precipitation Minimum temperature Soil drainage class
81% -‐ 90% > 90%
Probability of Occupancy for Current Conditions Probability of Occupancy 2 C increase
Brook Trout Probability of Occupancy < 10% 11% -‐ 20% 21% -‐ 30% 31% -‐ 40% 41% -‐ 50% 51% -‐ 60% 61% -‐ 70% 71% -‐ 80%
Model drivers Drainage area Forest cover Stream slope
Annual precipitation Minimum temperature Soil drainage class
81% -‐ 90% > 90%
Probability of Occupancy 4 C increase
Probability of Occupancy for Current Conditions
Resilience of occupancy to temperature increase
Brook Trout Probability of Occupancy < 10% 11% -‐ 20% 21% -‐ 30% 31% -‐ 40% 41% -‐ 50% 51% -‐ 60% 61% -‐ 70% 71% -‐ 80%
Model drivers Drainage area Forest cover Stream slope
Annual precipitation Minimum temperature Soil drainage class
81% -‐ 90% > 90%
Brook Trout Resilience Increase tolerated (°C) 0° 0.1° -‐ 0.5° 0.6° -‐ 1° 1.1° -‐ 1.5° 1.6° -‐ 2° 2.1° -‐ 2.5° 2.6° -‐ 3° 3.1° -‐ 8 .5°
Currently below threshold
Bringing it together Variable
Flow
Temperature
Season
Model Single-‐site demographic
Mul+ple-‐site demographic
Occupancy
Fall
↑ **
↑ ***
Winter
↓ **
↓ **
Spring
↔
↔
Precip ↑
Summer
↑ ***
NA
Fall
↓ **
NA
Winter
↔
NA
Spring
↔
↔
Summer
↓ ***
NA
Temperature ↓
Summary è
Congruent environmental effects on popula+on growth across scales
Increases confidence in generality of results ¤ Nega+ve effects of temperature ¤ Posi+ve effects of flow in fall and summer, nega+ve effects in winter ¤
è
Many brook trout popula+ons at risk in future Flow and temperature ¤ Extreme events ¤
è
Can iden+fy resilient popula+ons
Steve Hurley
Web app è
Map viewer ¤ Standard layers ¤ Data ¤ Model results ¤ Select a basin for scenario tester
è
Scenario tester ¤ Climate -‐> Landuse -‐> Environment -‐> Popula+on
response ¤ Evaluate management ac+ons under alternate futures hQp://felek.cns.umass.edu:8080/geoserver/www/data.html
Data types Sensi+vity of annual survival
