statement
Ecological Restoration Genetics
Summary
We have observed fine scale population structure in the annual selfing plant Arabidopsis thaliana as local and patchy with phenotypic variation due to new mutations and new allelic combinations. Whole genome association mapping is being used successfully to identify causative loci that can be confirmed by F2 crosses in the field. Strong selection in a heavily disturbed and fragmented landscape, leads to population explosion and collapse, however genetic variation often remains. Restricted gene flow in isolated populations, and a changing environment, may have similar destabilizing consequences on the distribution of genetic variation for many species occupying the same habitat. New sequencing technologies and the low cost of genotyping allows emerging model organisms to be developed such that model plant communities will be characterized at the genetic level. Thus, studies of the genetics of adaptation to environmental change can be compared across species that together shape their biotic environment. We have begun to seed multi year genetic studies on the interspecies successional process, priming ecosystem evolution. Remote sensing and population genomics are used to determine: How natural genetic variation across species shaped by internal biotic and external abiotic forces during plant community establishment and adaptation? This work should enable breeding for natural selection to restore biodiversity and ecosystem integrity as we adapt to global climate change.
Introduction: Where we are now
Ecosystem Health
In recent years it has become undeniably clear that ecological health of the environment is interconnected with human heath and well being, from the local to the global scale. For example air and water pollution causes disease and devastation of all species. At the cellular level, all biology is related (genes, cell structures, and signaling pathways, etc), but we are just beginning to elucidate the specific connections across the organism, community, and ecosystem scales. Photosynthesis is the foundation of our green planet harnessing sunlight to fix organic carbon and provide for the biosphere. Ecological interactions with host and microflora (pathogen, symbion, endo/ epiphyte) are recognized as essential for health and disease prevention, across kingdoms. We need detailed examples outside laboratory organisms and the clinic which span variable and evolving conditions of the real world. We need to expand our molecular toolkits to species that have different keystone positions and life history strategies across the tree of life and larger landscape. We are at an unprecedented time with our abilities to collect data, from sequences of molecules to development and behavior at scale. Current levels of data storage and computational power enable tremendous modeling complexity that suggest key testable field experiments. However time to resolve the crucial inter-dependencies of our position in, and our impact on, the biosphere is short. We must make important but flexible decisions to go beyond our traditional view of what causes health and disease and take a Systems Biology, or Systems Ecology view toward health of the biosphere.
A narrow view of environmental “health” focuses only on treating disease cased by pollution. With disease prevention or health promotion as our goal we must look to a broader view. How can natural systems be effectively restored and enhanced to provide holistic environmental health benefits? What combination of genotypes and species can establish and flourish, to restore fragmented, degraded, and/or contaminated sites? Traditional or molecular plant breeding has been extremely successful when specific traits and environmental conditions are known. When environmental health is broadly defined as the total life sustaining and environmentally resilient services provided by nature, a new form of "breeding" is required that restores genetic and species level biodiversity into connected functional communities within human ecosystems.
The current path has seen an increased withdrawal rate from Nature's EcoBank. Indeed humanity has been living on credit for at least the past couple decades. The amount of land required to produce the food, fiber, and fuel, and to absorb the wastes (including CO2) generated in their processing, exceeds the amount to biologically active land on the planet. This is called overshoot, the ecological footprint of humanity is between 1.3 and 1.5 Earths (http://footprintnetwork.org) depending on whether how much allot to all other species (conservation efforts are shooting for up to 10%). Humans now effect planetary change, as only seen before in the geological record. We have entered the Anthropocene. For example industrial process fix more nitrogen then all nonhuman process (bacterial) combined, mainly for agriculture which now leads to more erosion than all nonhuman processes combined. The last of our forests, fisheries, and fertile lands are on the table as demand grows exponentially. So human life depends on, and is a part of, natural biological processes that must be restored quickly to come back into a sustaining balance. Fortunately biological processes can effect change on a geological scale positively or negatively. Carbon and nitrogen sequestration, water filtration and storage, erosion prevention, and soil fertility generation, are but a few ecosystem services that plant and animal species use to create their own habitat. By restoring the primary biotrophic level, plants will fix and sequester the carbon dioxide we and our 4 wheeled pets breath. Vegetation and soils together contain about three times as much carbon as the atmosphere (Hawkins, 2008). Soil fertility, natural air and water filtering systems, and livable habitat is regenerated through an ecological process known as succession. First grasslands are established, followed by forests, all powered by sunlight. The conservation of species, communities, and intact ecosystems is necessary to seed the restorative processes.
Global Warming effects on plants
Most species have only a few alternatives in the face of climate change. They can migrate, adapt, or become extinct. Different species will have different fates. “Disruption in the synchrony between plants and pollinators is already affecting food security, nutrition and agriculture, as well as resulting in a decline of the numbers of pollinators themselves” (Hawkins, 2008). Population variation in flowering time for example would buffer against changes in pollinator migration and or provide opportunity for new pollinator/plant mutualisms to evolve. Bioclimatic modeling looks at environmental variables across the current species range, then predicts the future range under global warming scenarios. 25-42% of plant species will become vulnerable and likely to go extinct as their imputed range disappears (Hawkins, 2008). Both assisted migration and adaptation are needed which depends on population connectedness and genetic variability. The clock is ticking, however, for arctic and alpine species.
The Borevitzlab is extending our survey of genetic variation into native ecosystems beginning with perennial grasses. As new sequencing technology comes on line, we are scaling up to ecosystem level conservation genetics, to identify rare endemic subtypes and emerging invasives. Although the genetics and breeding are yet to be started, one popular application of mixed native species is prairie hay biofuel (Tilman , 2006). A comprehensive look at various ecosystem services, reveals that water and carbon sequestration (soil building) are crucial indirect effects promoting human health. Our work aims to 1) develop genomic data for several foundation perennial species in order to survey genetic diversity across a broad geographic and ecologial range. 2) Banked genetic diversity will be used for landscape level restoration or reconstruction. Environmental variables, ecosystem services, and evolutionary trajectories will be monitored as natural selection begins anew in a high diversity background. This will be experimentally monitored across distributed experimental diversity gardens (Foral Report Card NSF) and in large scale habitat reconstructions and natural area restorations.
Current and Previous Work
Arabidopsis thaliana is a small annual plant that is largely self pollinating and colonizes regularly disturbed environments. Cellular and development signaling pathways are largely known while responses to specific environmental stimuli in the lab can often be predicted. Recently attention has turned to ecological studies, investigating how populations respond to real world environmental conditions. Adaptation experiments test how genetic differences lead to differential reproduction or survival among various local conditions. In general the local conditions are defined by climate - the seasonally predictable moisture, temperature and light regimes in the area, as well as soil, shade and other biotic parameters including perennial plant competition. The evolutionary trajectory of a population, depends on founding diversity, in-migration, and new mutations. We have observed >5% effective outcrossing rates where multiple genotypes have shuffled variation which may out compete the parents. We have also observed new mutations in clonal fields where identical varieties differ significantly in flowering time. Both evolutionary mechanisms are at play, however reshuffling of existing genetic diversity, when available, can quickly generate phenotypic variation, enabling rapid adaptation.
My lab has focused on seasonal adaptation between specific geographic locations. Custom software controls modified incubators which simulate idealized seasons. For a given location, day length, temperature, humidity, light levels, and color mixture, changes throughout the day and year. We have measured responses among a diverse collection of accessions and in mapping populations (Liet al, 2006) to northern and southern simulated seasonal conditions. Major loci were identified showing genotype by environment interaction (local adaptation) and/or epistasis.
Traditional evolutionary models posit an adaptive landscape where each mutation or new gene combination as a step in a walk toward a fitness optimum. Ultimately the most highly adapted strain predominates. If the environment, including multi-species background, is changing, then there is no simple walk toward a narrow specialist optimum. Genetic diversity itself contains the toolkit to survive in a fluctuating ecosystem, while a narrow specialist genetic base is vulnerable to collapse and local extinction.
We have nearly completed the largest study to date of fine scale population structure in any species. By genotyping a hierarchical sample of ~6000 wild collections (Figure 3), we have mapped genetic variation at the local, regional, ecological, and continental scales. Many fields are clonal, especially in North America which was colonized in the past few hundred years. However genetic diversity warmspots and hotspots exist at several sites. They contains over 60 different and divergent haplotypes. These populations have maintained variation and re-assorted variation for decades. We will monitor population levels and diversity, among these evolutionary field replicates, documenting adaptation and/or maintenance of genetic diversity .
Arabidopsis thaliana is just a single species or data point on an ecological landscape. We have now developed Aquilegia as an emerging model organism (Abzhanov, 2008). Aquilegia was chosen as a text book example of adaptive radiations (Schluter 2000). This genus has evolved into more than 70 species during the last 10,000 years. In collaboration with Scott Hodge (UCSB), my lab has developed genetic and genomic tools, including genome sequence (in progress at JGI), a BAC based physical map, 85,000 ESTs, and >16,000 SNPs. Microarrays were designed and used to profile the 5 Aquilegia floral whorls for comparative evolution of the floral transcriptome (Voelckel et al, submitted). Natural populations (Figure 2A), mapping lines, and inbred horticultural accessions, are being phenotyped and genotyped (Figure 2B) for QTL mapping. One of the most spectacular stories in Aquilegia is the repeated evolution of floral pollinator preference. The ancestral floral type is small, blue, and bumble bee pollinated. This evolved twice into medium length, red, humming bird pollinated species, which then evolved at least four times into long, yellow, hawk moth pollinated flowering species (Whittall, 2007). Thus multiple independent speciation events were driven by plant pollinator preference and subsequent habitat differentiation. 1000s of wild strains of Aquilegia have been recently collected from Indian Dunes for population genetic analysis (Figure 4).
We are now expanding from model organisms to model ecosystems. Fermi labs has established the oldest, large scale native prairie reconstruction experiments in the country. Prairie ecosystems contain hundreds of species per square meter. Their biodiversity is on par with tropical rainforests; however more than two thirds of the species, and biomass is below ground. Life in the midcontinent has adapted to harsh winter (-40C)/ summer (+40C) temperature transitions, extreme floods and droughts, and fire conditions, by moving mostly underground. Globally, soils contain ~2/3 of all terrestrial carbon. Temperate grasslands contain more carbon biomass below ground than rainforests do above and below ground (Hawkins, 2008). Restoring these rangelands with their deep rooted grasses highlights a major untapped carbon sink. It takes 100s of years to remake the sod which early settlers built homes from but large gains are made in early years. In addition, perennial grasslands absorb water runoff, preventing soil erosion (eg, the dust bowl), while recharging ancient underground aquifers (eg Ogallala with 2X larger area then the great lakes). Work at the Fermi lab in 2008, is testing different stains of switchgrass, alone or in combination, and with other mixed species of tallgrasses. A 20 species prairie matrix plot has also been included as a control. The study will monitor above and below ground biomass, net carbon fixation, and ground water recharge, over a three year period. Biomass for cellulosic biofuel or pellet fuel bioenergy yield will also be determined.
Proposed work: Ecological Restoration Genetics
The proposed research program can be divided into two parts. First is the development of genomic tools and the assessment of existing genetic diversity for target native species across the region. Key divergent genotypes will be amplified in diversity gardens and seeds distributed. Second ecological restoration field sites will be established to enable breeding for natural selection. These field sites will be monitored at high resolution for environmental variables and evolutionary genetics.
Conservation Genetics: Identifying and preserving ancestral variation
Sequence Data/ Marker Development
For several understudied native prairie species, we will develop sequence variation data using second generation reduced representation sequencing (eg RAD markers) (~$20/sample 10-100k SNPs). Only a few hundred markers are needed to determine family pedigree structure, while ~10,000 are needed to resolve deep ancestry, and hundreds of thousands to millions are needed for genetic association mapping. These no longer require different genotyping technologies nor previously ascertained makers avoiding the traditional bias when markers are discovered in one population and typed on another.
In Arabidopsis close family members exist within a field, but have also recently spread around the globe with agriculture, removing almost all trace of genetic isolation by distance. In Aquilegia we tested horticultural and local collections and observed family structure among local populations and breeding lines. If this pattern is general, simple bulking of seeds of a given species from a particular location will represent an extreme bottleneck in genetic diversity. In addition, a particular local collection may not contain the diversity needed to reestablish in altered landscapes and or in future altered climates.
Target species:
We will select several foundation grassland species for genetic diversity mapping and conservation. They fall into 4 broad categories of cool season C3 perennial grasses, warm season perennial grasses, legumes, and sunflowers. Population samples will include all common horticultural varieties, arboretum and botanical garden collections, USDA native plant collections, as well as targeted collections in remnant natural areas. Initial sampling will reveal genetic diversity hotspots that can be revisited to collect additional samples. This will enable the selection of a species wide high diversity collection of ~384 lines that will be grown in various diversity gardens at collaborating locations.
Restoration Genetics
The second phase of the proposed work will establish reconstruction experiments at a network of marginal or abandoned agricultural sites. Partner organizations will have early access to diversity seed collections but we will make seed publicly available. Long Term Ecological Research Sites (LTRE) and National Ecological Observatory Networks (NEON) sites are available. Annually, after establishment, genetic variation will be tested along a transect testing 96 individuals for each of 12 species per site. The establishment phase is the most important stage at which natural selection will shape genetic variation. Selection however may not have a consistent direction as the plant community matures. In some sites reseeding of the genetic diversity seed matrix will be performed to reset the evolutionary clock and allow different species and genotypes to be selected in the background of others. Those unfit types will not be major contributors to the final composition. We expect selection to be local and different across years so predicting fitness will be especially difficult. Intercrossing will enable new varieties to be formed enabling selection to act on new gene combinations from wide crosses. If restorations are large enough, they may become sustainable on their own. Smaller locations can be reseeded every few years, and seeds could be transferred among small sites simulating habitat connectivity. Modest seed harvest can used to establish larger sites as seeds of diverse genetics will be chronically limited (especially for those varieties which put most of their energy into root biomass.
Real time environmental monitoring
Several remote sensing technologies will be used and data will be streamed back to a public web server. Remote time lapse cameras will be powered by solar panels. Air temperature, humidity, light intensity, and wind speed and direction will be recorded at 30 second intervals as will soil temperature and moisture levels. Air quality including CO2, O3, and other pollutants will also be monitored. The water table (level and quality) will be recorded electronically to determine the precise ground water recharge and nitrogen sequestration rates. Remote ecological studies are being performed at NEON sites however we envision a network of micro NEON sites as cost of remote sensing drops rapidly. The realtime ecological data describes landscape level behavior. This is a composite measure of whole grassland physiology and development (see http://hpwren.ucsd.edu as one example of remote ecosensing).
Teaching and outreach
Our data sets will be used for education. Virtual Ecological Observatories (VEcOs) are windows into nature to recruit Citizen Scientists (http://budburst.org) to document seasonal change and climate adaptation. Sites will host an annualBioblitz (public biodiversity survey). Finally real time data will be available to natural capital economists quantifying ecosystem service performance, and todaytraders of carbon, nitrogen, and water credits such that externalities on the environment can be internalized.
References
Abzhanov A, Extavour CG, Groover A, Hodges SA, Hoekstra HE, Kramer EM, Monteiro A Are we there yet? Tracking the development of new model systems. Trends Genet. 2008 May 30.
Li Y, Dunning M, Yu H, Bergelson, J, Nordborg M, Borevitz J, (submitted) The Scale of Population Structure in Arabidopsis thaliana
Li Y, Roycewicz P, Smith E, Borevitz JO (2006) Genetics of local adaptation in the laboratory: flowering time quantitative trait loci under geographic and seasonal conditions in Arabidopsis.PLoS ONE. Dec 27;1:e105.
Hawkins B, Sharrock S,Havens K (2008) Plants and climate change: which future? Botanic Gardens Conservation International, Richmond, UK
Savage C. (2004) Prairie: A Natural History Greystone Books
Schluter D (2000) The Ecology of Adaptive Radiation, Oxford University Press
Tilman D, Hill J, Lehman C. 2006. Carbon-negative biofuels from low-input high-diversity grassland biomass. Science. 314:1598-600
Voelckel C, Rensink W, Zhang X, Noutsos C, Hodges SA, Borevitz JO (submitted) Comparative Floral Transcriptomics
Whittall JB, Hodges SA (2007) Pollinator shifts drive increasingly long nectar spurs in columbine flowers. Nature. Jun 7;447(7145):706-9
