finding balance: considering species' traits, species distribution … · 2016. 11. 18. ·...

Finding Balance: Considering Species' Traits, Species Distribution Models, and

Climate Forecasting in Seed Sourcing Decisions

Arlee M. MontalvoNov. 15, 2016 Do No Harm Workshop, Davis, CA

Acknowledgements:

My collaborators!Erin C. Riordan- the modeling and mapsJan L. Beyers- editing, profile assistance

Thanks to USFS Pacific Southwest Research Station and RCRCD for funding and UCR and UCLA for

logistical support!

California shrublands occupy a diverse landscape:Ecological Sections and Subsections Goudey and Smith (1994) updated with ECOMAP (2007)

Parent geology

Precipitation and temperature

topography ocean influence

continental influence

Elevation: 300 to 11,500 ft Precipitation: 6 to 40 inches Temperature: 40° to 70°F Growing Season: 150 to 300 daysGoudey & Smith 1994

Ecological Sections can be diverse and contain many contrasting Subsections

alluvial scrubchaparral coastal sage scrub

California shrublands and their foundation plant taxa are diverse

Environmental diversity supports amazing biological diversity

protecting, supporting, managing native biodiversity

ALSC CHAP CSS MIX

Acmispon glaber var. brevialatus † suffr subshr X XAcmispon glaber var. glaber † suffr subshr X X X XAdenostoma fasciculatum shrub X X XArctostaphylos glandulosa (3 subsp) shrub XArtemisia californica subshrub X X XCeanothus crassifolius var. c. shrub X XCeanothus cuneatus var. cuneatus shrub XCeanothus leucodermis shrub XCeanothus megacarpus var. m. shrub XCeanothus oliganthus shrub XCeanothus perplexans shrub XCeanothus tomentosus shrub XCeanothus vestitus shrub X

Cercocarpus betuloides var. b. shrub X XCorethrogyne filaginifolia† suffru per. X X X XEncelia californica subshrub XEncelia farinosa subshrub XEriodictyon crassifolium var. c. suffr subshr X XE. c. var. nigrescens suffr subshr X X X

Scientific Name Life Form‡Plant Community

ALSC CHAP CSS MIX

Eriodictyon trichocalyx var. lanatum suffr subshr X XE. t. var. trichocalyx suffr subshr X X XEriogonum fasciculatum var. fasciculatum subshrub X XE. f. var. foliolosum subshrub X X X XE. f. var. polifolium subshrub X XEriophyllum confertiflorum var. c.† suffru per. X X XHesperoyucca whipplei suffr rosette X X X XHeteromeles arbutifolia shrub X XLepidospartum squamatum shrub XMalacothamnus fasciculatus var. f.† subshrub X X XMalosma laurina shrub X X X XPrunus ilicifolia subsp. ilicifolia shrub X XQuercus berberidifolia shrub X XRhamnus crocea shrub X XRhamnus ilicifolia shrub XRhus ovata shrub X XSalvia apiana suffr subshr X X XSalvia mellifera subsh/shrub X X X X

Scientific Name Life Form‡Plant Community

Many foundation species of shrubs are distributed across multiple Ecological Sections

and plant communities

For each of 36 shrubs & subshrubs:

• Assembling profiles with information about physiology, ecology, demographics, life-history traits, and population genetics to inform interpretation and practical applications of model results and seed sourcing decisions

• Running models that estimate current and future suitable habitat

Most taxa have been studied for fire response, many for drought tolerance, but few for genetic patterns

Pratt et al. 2014. Oikos. 123(8) pp 953-963. Figure 3.DOI: 10.1111/oik.01156

Genetically based latitudinal variation in Artemisia californica secondary chemistry

Common garden studies have shown correlated patterns in use of plants by arthropods and in traits related to water economy

Distribution of California sagebrush (Artemisia californica)

Acmispon glaber (aka Lotus scoparius), California broom, deerweed

•self-compatible subshrub with infra-specific variation

A. g. var. glaber A. g. var. brevialatus

(Montalvo & Ellstrand. 2000 Conservation Biology, 2001 AJB)

•differ in morphology and habitat affinities•two named varieties - distributions differ but overlap

•common garden work local adaptation, outbreeding issues

Careful choice of species, subspecific taxa and seed sources for projects can:

• Minimize risk of maladaptation• Maintain variation and adaptive potential• Reduce risk of inbreeding and outbreeding

depression• Preserve important interactions• Increase long-term success of projects

How do we factor in climate change?Observed shifts in climate variables for 1981-2010 relative to 1951-1980

using CA-BCM downscaled climate data (270 m resolution)

Δ MIN DJF Temp (C) ΔMAX JJA Temp (C)

Δ PPT % Δ CWD (mm)

Dealing with a complex world: Outline

• Starting point for seed sourcing• Consider climate change and Species

Distribution Modeling • Plant traits important in decision frameworks• Decision frameworks and provenancing

models to help keep us on right track

A. D. Bower, J. B. St. Clair, V. Erikson. 2014. Ecological Applications 24:913-919. Ecological Archives A024-053-A1. Based on high resolution climate data & aridity index.

a starting tool to be combined with expert knowledge of plants

For the many species with no seed transfer research, Provisional seed zones for native plants reflect the

complex landscape and associated climate

Seed Zones Mobilehttp://www.fs.fed.us/wwetac/threat_map/seed_zones/Seed_Zone_Google_Map_Links.pdf

How effective are these tools for southern CA?

What about climate change?

Will rate of climate change exceed species’ capacity to respond?

Prov. Seed Zones

35 - 40 Deg. F. / 6 - 12Tmin/AHM

CNRM RCP 8.52040-2069

MIROC RCP8.52040-2069

Baseline1951-1980

Genetic maladaptation of coastal Douglas-fir seedlings to future climates

St. Clair, B. J., and G. T. Howe. 2007. Global Change Biology 13(7): 1441-1454

Assisted migration: ”the purposeful movement of

individuals or propagules of a species to facilitate or mimic natural range expansion or

long distance gene flow within the current range, as a direct management response

to climate change” (Havens et al. 2015. NAJ)

Detailed studies of long-lived tree species

Seeding sourcing for the future?

• First, what is projected loss of suitable habitat?• Consider risks from moving too much or too soon:

• poor adaptation to current conditions • growth phase mismatches with seed/pollen dispersers• outbreeding depression• unanticipated changes in community interactions• unanticipated aggressiveness/weakness in novel environment

• Are there other, interacting risk factors?• Action should not cause more harm than no action

Dealing with a complex world




• High resolution (270 m) baseline (1951-1980), current (1981-2010) and projected (2049-2060) climate and hydrologic surfaces

• Variables used in modeling: strong drivers of plant distribution– Winter Tmin, Summer Tmax, – T seasonality– Winter PPT, Summer PPT– Climatic Water Deficit (CWD), Actual

Evapotranspiration (AET)

Downscaled climate data: CA-BCM

California Basin Characterization Model (CA-BCM): applies amonthly regional water-balance model to simulate hydrologic responses to climate at the spatial resolution of a 270 m grid

Flint et al. 2013 Ecological Processes

Projected change in southwestern CA climate 2040-2069 relative to 1951-1980

Increase in minimum winter Temp (°C)

Perc

ent c

hang

e in

ann

ual P

PT (%

)

RCP

Projected change in southwestern CA climate 2040-2069 relative to 1951-1980

Increase in minimum winter Temp (°C)

Perc

ent c

hang

e in

ann

ual P

PT (%

)

“Hot/Wetter”

“Hot/Drier”

“Hottest/Very Dry”

“Warmer/Drier”RCP

“Hot/Drier”

Future conditions(2040‒2069)

future

Planning Tools

Habitat suitabilitybaseline+

Baseline conditions(1951‒1980)

Species occurrences(Herbarium records

and plot data)

MAXENTSpecies Distribution

Model (SDM) algorithm

Tseas

Species distribution models for southern California region

• Estimates spatial pattern of suitable habitat and shifts in suitable habitat by pooling data over species’ range

• Results vary with global change model employed• Some caveats (concerns)

– Doesn’t show population level patterns of variation, biotic interactions, or demographics

– Not all populations likely to do well in all suitable habitat– Assumes no adaptive capacity to respond to change– Other, very important risks factors not in model– Climate may interact with other drivers of change

Extending SDM to incorporate population level data is enormously complex

Species Distribution Modeling (SDM)

SDM’s run on the southern California extent of species’ ranges and on infra-taxa

• Most species likely to have populations that differ genetically from north to south

• SDM results on smaller extent differ from SDMs run on full species range

• Resulting projections of habitat suitability are likely more realistic for the extent under consideration

Pratt et al. 2014. Oikos. 123(8) pp 953-963. Figure 3.DOI: 10.1111/oik.01156

Genetically based latitudinal variation in Artemisia californica secondary chemistry

Species Distribution Models (SDM) estimate future stability, loss, and gain of suitable habitat

SDM of projected change in habitat suitability for Artemisia californica

=urban/agriculture

Baseline (1951-1980) suitability Projected (2040-2060) stable suitabilityOccurrence data

modeling extent

MAXENTalgorithm

Projected future suitability lossProjected future suitability gain

Example of plant with infra-specific taxa

•distributions of Acmipson glaber varieties differ, but they overlap•common garden work showed local adaptation and outbreeding issues

A. g. var. glaber

A. g.var. brevialatus

Acmispon glaber var. brevialatus Acmispon g. var. glaber

Baseline (1951-1980) habitat suitability

Acmispon glaber var. brevialatus Acmispon glaber var. glaber

SDM projected habitat suitabilityto mid century(2040-2069):

var. brevialatus,55–98 % stableloss > gain (3)

var. glaber, 1–65 % stable; loss > gain (5)

= high climate exposure

loss

gain

stable

loss

gain

stable

Species vary in their exposure, resilience, and vulnerability to climate change and translocation

• What traits capture differences in species’ ability to persist in place, adapt, and migrate?

• What traits are associated with different levels of risk of maladaptation or outbreeding depression?

To navigate decision frameworks

We need to know about the plants!

Plant traits influence migration (gene flow)Fruit type x primary / secondary dispersers

Gravity / ants, rodents, water Wind / ants

Birds, squirrels / rodents

Birds, mammals / rodents

near far(km)(meters)

Stature and structure influence gene flowWind dispersed seeds or pollen

far (100’s of meters)near (within 10 m)

height

Plants x animals x habitats influence migration

Type of pollinator: foraging distance, constancy

near far

gnats, beetles, tiny solitary bees butterflies, moths bees, hummingbirds,

bumblebees, social bees (honeybee)

(km)(meters)

Barriers to migration across a landscapeHeterogeneity and fragmentation can influence dispersal

Parts of California are highly fragmented, especially along coastDecreasing porosity

of barriers

Correlates between life-history traits and spatial genetic structure (differences among

populations) in plant species

Correlation with spatial genetic structure

Trait Highest LowestBreeding system Selfing species Outcrossing, wind-pollinated

Life form Annual Long-lived, woody perennial

Seed dispersal mechanism

Gravity Gravity then animal-attached

Successional status Early Late

Taxonomic status Dicots Gymnosperms

Regional distribution Temperate Boreal-temperate

from Rogers & Montalvo 2004, derived from Hamrick & Godt 1990

Higher levels of genetic differentiation and structure are associated with:

• Habitat heterogeneity and geographic isolation• Low gene-flow potential (e.g., short, self-compatible

herbs, gravity dispersed seeds) • Local adaptive differences and unique gene interactions• Local adaptation (home site advantage) at smaller scale• Higher likelihood of problems following mating with

other populations

On this end of continuum, one would decrease distance limits to collection

Modified from Havens, Vitt et al. 2015. Natural Areas Journal

More conservative/near distance sourcing

More relaxed/longer distance sourcing

Narrow and/or habitat specialistLittle long-distance gene flow

Low phenotypic plasticityNarrow environmental tolerance

Widely distributed/or generalistExtensive long-distance gene flow

High phenotypic plasticityWide environmental tolerance

Species traits

Habitat traits

In between?

Historically fragmentedHigh quality

Ancient/stable landscape

Recently fragmentedLow quality/degraded

Younger/dynamic landscape

Taxonomic uncertainty, cryptic speciesHigh hybridization potential

Low rates of evolution (conserved)

Taxonomic stability, well knownLow hybridization potential

High rates of evolution

Taxonomic understanding

A framework to aid “distance” decisions

Distance ecological/genetic/geographic

Modified from Havens, Vitt et al. 2015. Natural Areas Journal

More conservative/near distance sourcing

More relaxed/longer distance sourcing

Narrow and/or habitat specialistLittle long-distance gene flow

Low phenotypic plasticityNarrow environmental tolerance

Widely distributed/or generalistExtensive long-distance gene flow

High phenotypic plasticityWide environmental tolerance

Species traits

Habitat traits

In between?

Historically fragmentedHigh quality

Ancient/stable landscape

Recently fragmentedLow quality/degraded

Younger/dynamic landscape

Taxonomic uncertainty, cryptic speciesHigh hybridization potential

Low rates of evolution (conserved)

Taxonomic stability, well knownLow hybridization potential

High rates of evolution

Taxonomic understanding

How does Acmispon glaber var. brevialatus score?

Evaluate direction of shifts in future climate of ecoregions and seed zones and compare with shifts in habitat suitability revealed by SDMs

Can taxon likely tolerate expected shifts in place? (secure refugia)

Candidate for “assisted migration”. Consider scale of expected shifts in

habitat suitability relative to scale of local adaptation, migration & risk

Taxa with low levels of genetic diversity, low gene flow, or adapted to highly localized

climate or edaphic factors?

NO

YES

Is taxon able to migrate &/or is it sufficiently genetically diverse for adaptive evolutionary response?

If taxon SDM is available, is there substantial risk that habitat climatic suitability will decline?

NO

NO

YES

Limit zone combinations for seed sourcing to adjacent zones and logical

direction of expected shifts.

YES

No seed zone shifts warranted

YES

Restore migration corridors?

NO

Taxon with high levels of plasticity, genetic diversity, gene flow, or

historical climate tolerance

Modified from Shoo, Hoffmann, Garnett, et al. 2013. Climate Change 119:239–46

YES

Provenancing models deal with common questions about seed sourcing

• How can we move seeds, while minimizing risk of maladaptation and genetic mismatches?

• What type of seed sourcing (provenancing) model should we use within seed zones or among seed zones?

“Climate-adjusted provenancing”

Projected increase in aridity

(modified from Prober et al. 2105. Frontiers Ecol. Evol.)

planting site

Source populations

Summary

• SDM can be used with detailed ecological, physiological, and population genetic information to inform seed sourcing decisions.

• Not all baseline or future “suitable habitat” is appropriate or available.

• Distance models for sourcing seeds within current climates can also apply to assisted migration.

• Candidates for assisted migration: high climate exposure, low gene flow/adaptive capacity, or highly compromised dispersal capacity

• Other risk factors may be more important or interact with climate change

Managing biological diversityin southern California shrublands

alluvial scrubchaparral coastal sage scrub

Questions?

Building profiles for ~44 taxa, fully referenced

• Species: Photos, taxonomy, relationships• General: Mapped occurrences, life-history traits• Habitat: Associated vegetation and environmental attributes• Climate change and projected future suitable habitat: Species

distribution models and forecasts, fragmentation• Growth, reproduction, and dispersal: Including fire effects• Biological interactions: Competition, microorganisms, herbivory,

seed predation, animal dispersers• Ecological genetics: Ploidy, plasticity, pattern and scale of

genetic variation, translocation risks• Seeds: Dormancy, germination, planting, seed increase• Ecological and evolutionary considerations for restoration:

summary and implications

FIRST SET TO BE POSTED IN 2017 at www.RCRCD.ORG

National Native Seed Strategy for Rehabilitation and Restoration/ 2015-2020

• The vision: the right seed in the right place at the right time• The Mission: To ensure the availability of genetically

appropriate seed to restore viable and productive plant communities and sustainable ecosystems.

A public-private partnership of organizations that share the goal: to protect native plants by ensuring that native plant populations and their

communities are maintained, enhanced, and restored.

Protecting/supporting/managing biodiversity

finding balance: considering species' traits, species distribution … · 2016. 11. 18. ·...

Documents