Climate change is both an ecological and evolutionary event that can force assisted migration and genetics-based ecosystem engineering

Tom Whitham, Merriam-Powell Center for Environmental Research

Not Bob Marley Heat map photo of & by Tom Whitham SEGA site at The Arboretum at Flagstaff – A network of 10 common gardens along an elevation gradient to develop solutions to global change

Part 1 - How locally adapted are plants? Importance - The more locally adapted, the greater the impacts of climate change. (sega.nau.edu) $4.5 million NSF/NAU, GO, and NGO Participants: USFS, NPS, BLM, BOR, TNC, AZ Game & Fish, Babbitt Ranches, Grand Canyon Trust, & The Arboretum at Flagstaff

Illustration by Paul Heinrich

If plants must move to survive future climate conditions, how do we decide on which ecotypes and genotypes to use in restoration projects? SEGA network provides next generation genetics-based infrastructure to scientifically make such decisions. Geographical Distribution of Ecotypes Geographical Central California Ecotype adapted ecotypes have evolved in Utah High Plateau Ecotype response to environmental differences across the range of P. fremontii.

Sonoran Desert Ecotype

Structure Analysis based upon molecular marker

Ikeda et al. 2017 Global Change Biology Utah High Plateau Ecotype Central California Ecotype

Sonoran Desert Ecotype

Most Predictive Environmental Variables

Bio.15 – Precipitation seasonality Bio.6 - Max temp of warmest month Bio.11 – Mean temp coldest quarter

Using genetically informed ecological niche modeling (gENM) with Maxent, we found that the regions occupied by different ecotypes will shift with projected climate change and will diverge spatially even more than their current distributions (Ikeda et al. 2017 Global Change Biology). Genetics-based models are up to 12x better at predicting ecoregion test points

GeneticsGenetics NoNo Genetics 0

-2

-4

-6

-8

-10

-12 Binomial Probability (ln) Probability Binomial

Predictive accuracy Predictive -14

-16 Central Sonoran Utah -18 California Valley Desert High Plateau California Central Valley Sonoran Desert Utah High Plateau EcoregionEcoregion Ikeda et al. 2017 Global Change Biology Reciprocal common gardens show finer scale local adaptation within the Sonoran desert ecotype

Reciprocal TNC Dugout Ranch, Canyonlands Common Cooler Garden – MAT 10.7 °C Gardens

AZG&F Horseshoe Ranch Intermediate – MAT 17.2 °C

BLM Mittry Lake, Yuma Hot Garden – MAT 22.8 °C

Location map of 16 provenance collection sites (leaf icon) of Populus fremontii and the three common garden locations (leaf with circle). The central garden is also a collection site. The shading corresponds to the degree-days above 5°C (DD5) throughout the region: red represents high DD5, blue low DD5. (Cooper et al. unpub.) March 8, 2017 – Photos by Tom Whitham

Agua Fria Creek

Populus fremontii field trial

Indian Creek

4600 tree common garden on Arizona Game & Fish Dept. lands at Horseshoe Ranch surrounded by Agua Fria National Monument. Hot Garden Intermediate Cool Garden Long Growing Season Short Growing Season

Populations are locally adapted within an ecoregion Population level mean (+/- 1 SE) survival correlations with bud set date in each of the three common gardens. Populations are colored by the mean annual temperature (MAT °C) of their source provenance. In Yuma, survival is highest in the hotter source populations and is positively correlated with later bud set. The opposite is true in the coldest Canyonlands garden. From Cooper et al. 2018 Global Change Biology (in press). Fremont cottonwood P. fremontii

Geographic mosaic of divergent functional trait selection Divergent selection in quantitative traits affected by climate Qst-Fst of plant functional traits (bud phenology, height, DRC, SLA) across the three common gardens (Cooper et al. unpub. data). Same has been shown for narrowleaf cottonwood, P. angustifolia (Evans et al. 2016) Part II. Genetic differences in the functional traits of plants affect community structure and ecosystem processes. Importance – If climate selection is non-random, then community structure and ecosystem processes will change. Genetic “footprints” of trees can be large: The genetic links between terrestrial and aquatic ecosystems.

Intraspecific differences in cottonwoods affect stream macro-arthropod communities.

Populus angustifolia

Emergence trap photo by Zacchaeus Compson 1.0

0.8 996 1008 1012 0.6 1017 1020 0.4 RM2 T15 0.2

Axis 2 0.0

-0.2

-0.4

-0.6

-0.8 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Axis 1 Different genotypes of P. angustifolia support different stream macro-invertebrates. Compson 2016 Ecosphere Community Phenotypes

Traditional Phenotype Frequency of Resistance

Aphid survival AFLP genetic markers Ecosystem Phenotypes

Whitham 1989 Science, Dickson & Whitham 1996 Oecologia, Schweitzer et al. 2006 Oikos, Keith et al. 2010 Ecology, Zinkgraf et al. 2016 J. Insect Physiology Genetic Linkage Map of Populus lg1 lg6 QTL Mapping InsectLeafArchitecturePhenology ChemistryResistance

lg8

Salicortin lg18 lg12 Spring Leaf Flush Resistance to P. betae

Second Year Branch Number

Condensed Tannin

Woolbright et al. 2008 Heredity Matt Zinkgraf et al. 2016 J Insect Physiology Woolbright et al. 2018 Ecology & Evolution Illustration by Victor Leshyk

Why do different genotypes support different communities? Many functional traits result in individuals being very different from one another and affecting other traits such as diversity, stability, and network structure. San Francisco Peaks & Wupatki National Monument view from Navajo Nation Photo by Tom Whitham

Communities are far more linked than previously envisioned. Lichen 2010 P ! 0.05 Lichen 2010 0.05

Phyllosphere FungaLeaf Pla tLhogeeansf 2010 Ectomycorrhizal Journal of Ecology EM Fungi 2006 Trunk Pathogens 2010 Fungi 2006 Soil

Fungal Leaf Leaf Pathogens 2009 SoiSl oiBal Bacctteriria a2004 Pathogens 2009 2004

LeaLf eMaf Modiodififyiers ng2010 Soil Fungi Arthropods 2010 Soil Fungi 2004 2004 Populus

TwigEndophyt EEndophytesndophyees 2006 s angustifolia 2006

The network of correlatedLiLcihechen 2010 2010 communities is defined by individual tree Bgenotypes – the importance of maintaining network structure in restoration. FungaLeaf Pla tLhogeeansf 2010 Ectomycorrhizal Pathogens 2010 EFMungi Fungi 2006 2006 Community-genetic correlations - changes in the composition of one community among plant genotypes that are mirrored by changes in the composition of another community.

Fungal Leaf Leaf Pathogens 2009 SoiSl oiBal Bacctteriria a2004 Pathogens 2009 2004

LeaLfe aMf Modiodififyiers 2010ng Soil Fungi Soil Fungi 2004 Arthropods 2010 2004

TwEindophytg Endophyees 2006 s 2006 Fremont cottonwood on the Little Colorado River - Photo by Tom Whitham

Part III. Climate change creates a mis-match of once locally adapted plants and the new environment that affects plant survival, biodiversity, and mycorrhizal mutualisms. Climate change results in mis-matches with the local environment.

In just 25 years plant hardiness zones have shifted northward and upward in elevation by one zone. Photos by Hillary Cooper (top) & Tom Whitham (bottom) Cottonwood mortality on Bill Williams River National Wildlife Refuge – March 28, 2017

Stand die-off due to record drought – plants are no longer adapted to the local environment. As the local stock dies out, what populations and genotypes should be used in restoration? Record drought of 2002 – An evolutionary event that changed the genetic structure of the tree population.

Sthultz et al. 2009 Photo by Tom Whitham Global Change Biology Sept 15, 2004 - North of San Francisco Peaks, AZ Diverse arthropod community on pinyons negatively affected by drought 40 155 arthropod species

30 Within a site, increased 20 stress negatively affects 10 arthropod diversity.

Total richness per tree per richness Total 0 Species Richness Stress Index = (standardized r2=0.53, p<0.0001 branch dieback + 0.0 0.5 1.0 1.5 2.0 2.5 number of needle cohorts + 25 Stress Index radial trunk growth)

20

15

10

5

Total abundance per tree per abundance Total 0 Abundance r2=0.53, p<0.0001 -5 Stone et al. 2010 Oecologia 0.0 0.5 1.0 1.5 2.0 2.5 Stress Index Part IV - Solutions to the mis-match EVOLVE, ACQUIRE BETTER MUTUALISTS, MOVE or DIE

Painted Desert from SP Crater Photo by Tom Whitham Genetics-based responses to

Genetic Variation climate change in Plasticity

Plasticity

Genetic

A two-environment reaction norm showing the components of phenotypic variation of four genotypes: G = trait variation due to population genetics within a single environment, E = trait variation due to change in environment (plasticity), GxE = the variation in plasticity among genotypes. Phenotypic variation (VP) = VG + VE + VGxE. From Cooper et al. 2018 Global Change Biology (in press). Rapid Evolution with Drought Drought susceptible trees were >3X more likely to die during the 2002 record drought than drought resistant trees. 80 70 60 50 40 30

% mortality% 20 10 0 Drought Drought Susceptible Resistant Pinyons Pinyons Gehring et al. 2014 Molecular Ecology Seedling from Drought Tolerant Seedling Mother from Drought Intolerant Mother

A 2500 tree pinyon pine common garden shows that 1) drought tolerance is a heritable trait and 2) mycorrhizae on drought tolerant trees are better mutualists. Figure 3

Drought Mycorrhizal Tolerant 1. Mycorrhizal communities Community Seedlings are heritable plant traits.

Adults

Adults Drought Intolerant Seedlings

Figure 4 2. Drought tolerant Drought mycorrhizal communities are Drought Tolerant better mutualists. Intolerant

Gehring et al. 2017 PNAS Pinyon Seedling Performance Photo by Lloyd A. Whitham, circa 1950, nursery established 1863 Assisted gene flow/migration has been used inappropriately for a long time; we must get smarter in its use. Hot Garden Intermediate Cool Garden Long Growing Season Short Growing Season

Assisted Migration

Assisted Migration

No Assisted Migration Options

Common gardens allow us to predict the best source populations for a given level of climate change Cooper et al. 2018 Global Change Biology (in press). Phenotypic plasticity defines limits to assisted migration

Home - Keams Canyon, AZ

Chevelon Garden - ∆MAMT = 3.2 °C hotter transfer distance

Yuma Garden - ∆MAMT = 12.7 °C hotter transfer distance

Photos and data by Jackie Parker A fundamental issue in assisted migration is if you move plants to mitigate the impacts of climate change, will plants acquire the communities of their home sites? In other words, if you build it will they come?

Odgen Nature Center restoration site Photo by Tom Whitham Up to a point, if you built it they will come.

With transfers of 18 and 48 km, garden and wild trees support similar communities, but at 90 km they are quite different (Keith et al. unpublished data). Summary

Marble Canyon dust storm “haboob” Photo by Tom Whitham Part I. Plants can be very locally adapted. negative effects with climate change Part II. Plant genotype can define communities. negative effects with climate change Part III. Climate change creates a mis-match of once locally adapted plants and the new environment. negative effects with climate change Part IV. Solutions to the mis-match. passive - genetic variation & plasticity; human intervention - assisted gene migration & ecosystem engineering; inoculate with better mutualists Collaborators in Community Genetics and Genetics-Based Restoration Rachel Adams – plant ecology Gery Allan – molecular ecology Petter Axelsson – transgenic trees Joe Bailey – community ecology Randy Bangert – biogeography Rebecca Best – ecology & evolution Davis Blasini – ecophysiology Helen Bothwell – phylogeography Posy Busby – ecological plant pathology Abraham Cadmus – ecophysiology Aimée Classen – soil ecology Zacchaeus Compson – aquatic ecology Hillary Cooper – phylogenetics Sam Cushman – landscape genetics Steve DiFazio – molecular ecology Rodolfo Dirzo – community ecology Luke Evans – population ecology Sharon Ferrier – conservation ecology Dylan Fischer – ecophysiology Paul Flikkema – systems engineering Kevin Floate – insect ecology Catherine Gehring – microbial ecology Heather Gillette – molecular ecology Kevin Grady – restoration Steve Hart – ecosystem/soil ecology Erika Hersch – ecological genetics Joakim Hjältén – ecology Lisa Holeski – genetics & chemistry Kevin Hultine – invasive species Dana Ikeda – climate modeling Julia Hull – endophytes Nathalie Isabel – molecular ecology Karl Jarvis – phylogeny Art Keith – insect community ecology George Koch – ecophysiology Tom Kolb – plant physiology Lela Andrews - molecular ecology Jamie Lamit – microbial ecology Matthew Lau – network modeling Carri LeRoy – aquatic ecology Rick Lindroth – chemical ecology Jane Marks – aquatic ecology Lisa Markovchick – microbial ecology Tamara Max – molecular ecology Richard Michalet - facilitation & ecology Nashelly Meneses – ecological genetics George Newcombe – plant pathology Emily Palmquist – hydrology Jackie Parker – plant ecology Brad Potts – quantitative genetics Jen Schweitzer – ecosystems David Smith – landscape ecology Steve Shuster – theoretical genetics Chris Sthultz – plant ecology Amy Whipple – ecological genetics Tom Whitham – community ecology Gina Wimp – community ecology Todd Wojtowicz – litter arthropods Troy Wood – ecology Scott Woolbright - molecular genetics Adam Wymore – aquatic ecology Matt Zinkgraf – molecular genetics

GO & NGO collaborators: Mary McKinley – Ogden Nature Center, Gregg Garnett – Bureau of Reclamation, Kris Haskins – The Arboretum at Flagstaff, Paul Burnett – Utah Dept. of Natural Resources, Billy Cordasco – Babbitt Ranches Outreach – Lara Schmit, Victor Leshyk - NAU

Macrosystems MRI