Landscape Genomics and Forest Health Network Overview

Knowledge of the adaptive genetic potential of forest tree populations is fundamentally and critically important for evaluating their vulnerability to climate change. Thriving forests are key to sequestering carbon and consequently to mitigating the impacts of climate change. In this multidisciplinary project, the overarching goal will be to identify “all” genes under selection from geoclimatic factors, determine their allelic diversity, and seek approaches to applying this knowledge in land-management decisions influencing genetic resource allocation. We will begin by using modern landscape genomic approaches to obtain a comprehensive understanding of the standing adaptive genetic potential of eight important conifer species of the California forest ecosystem. We have chosen the California forest ecosystem because of its high species diversity and ecological complexity, rich vegetation history databases, and proximity to UC Davis. Next we will combine understanding of the patterns in this adaptive genetic landscape with detailed spatial databases of historic and contemporary climate and vegetation to infer the processes that created the genetic patterns. We will use these insights together with projected future climate data to predict future distribution of vegetation types and genetic diversity and evaluate approaches to maintain and make better use of conifer genetic diversity. To maximize the translational aspects of our research, we will develop a database and web-based query and geographic mapping tools that can be used by forestland managers to incorporate information about genetic adaptation into their reforestation, restoration, and conservation plans. Finally to assure broad usage of this resource and to incorporate user feedback, we will develop a comprehensive Research and Education Integration Program to teach land managers to use the newly developed tools. A key and novel component of our Research and Education Integration Program is that we have assembled an advisory committee of lead state and national forestland managers who have contributed to the project design and will continue to advise the basic research and discovery throughout the project.

The relevance and importance of forest trees

Forest trees are the dominant life form on roughly 30% of the earth’s surface (3.87 billion ha; FAO 2001). Approximately one third of California’s 99.6 million acres is forested (Christensen et al. 2008; including virtually the entire Sierra Nevada and Coastal mountain ranges. Conifers grow from sea level, in the Coast Range, to 12,000 feet in the Sierra Nevada. California conifer forest ecosystems are the most diverse in the nation, or for that matter, the world, and contain 14 of the 16 conifer genera in the USA. California is home to over 50 conifer species (Lanner 1999). The many ecosystem functions of forests in California, which include carbon storage and water production for the most populous state in the nation, are driven by forest cover. In short, forests are key to the economic and ecologic vitality of the region.

California forests are ecologically complex, harboring a wide variety of edaphic and climatic conditions. Many of the conifers in California forests exhibit relatively broad natural ranges while others exist as marginal endemics. For most species, research has demonstrated that local populations are most adapted to existing environmental conditions (Rehfeldt et al. 1999, St.Clair et al. 2005)). However, as climatic conditions change, populations of some species are becoming suboptimally adapted to some or most of their current range. These species and populations are clearly at risk. Fortunately, conifers, in general, harbor enormous genetic diversity, including adaptive genetic diversity, which might allow adaptation to changing environments. Unfortunately, there is a near complete lack of understanding of the standing adaptive potential of most forest tree populations. The overarching goal of this project is to develop a deep understanding of the adaptive genetic potential of forest tree populations and design strategies to help apply this knowledge to forest management strategies that will help to mitigate the effects of climate change.

Landscape genomics is an emerging new discipline

The discipline of landscape genomics combines population genetics and landscape ecology to study patterns of demographic and adaptive genetic variation across heterogeneous landscapes. Research outputs can be used directly by natural resource (plants and wildlife) managers to monitor, restore, and conserve threatened or endangered species or populations. To date, there have been few large-scale applications of this approach to natural plant and animal populations. Forest tree species, with broad contiguous populations, yet often with local adaptation, are excellent models for empirical study in landscape genomics. The study proposed here responds directly to the Genome-Enabled Plant Research (GEPR) section of the Plant Genome Research Program: “proposals are especially encouraged in the following areas, 1. Plant responses to environmental stresses, especially as they relate to climate change”. Furthermore, this proposal also responds to the Life in Transition (LiT) activity; “the Directorate for Biological Sciences at NSF encourage submission of proposals that address the biochemical, molecular, cellular genetic and/or organismal underpinnings of adaptation and biological feedbacks to climate change”.

Forest geneticists in the US have studied the genetic basis of adaptation in forest trees for centuries

Forest geneticists have employed the common garden techniques, popularized by Jens Clausen, David Keck, and William Hiesey, for centuries (Stettler 2009). Such studies were compelled by the failure of forest tree plantings using nonadapted sources. Trees, unlike most agriculture crops, are grown without any agricultural inputs (irrigation, fertilizer, weed and pest control), and must be well adapted to the sites at which they are grown. The genetic source of the seeds used to produce planting stock for reforestation following harvest or disturbance is critically important in determining the growth and survival of plantations. The guiding principle for selecting seed sources has consequently been “local is best”. Defining what is “local” has been a major focus of forest genetics research for many decades.

The primary outputs from these common garden experiments are the delineation of “seed zones” or “breeding zones” which guide subsequent regeneration efforts. This approach provides high quality information but takes years and great expense to obtain. Forest geneticists are keen to develop approaches that could reduce the time and cost to obtain information on the patterns of adaptive genetic variation in tree populations. Genetic marker information is an obvious alternative to common gardens, however the suite of marker tools (isozymes, RAPDs, RFLPs, AFLPs, SSRs) has generally revealed patterns of demographic, not adaptive, genetic variation and thus have been of little utility to forestland managers (Gonzalez-Martinez et al. 2006).

High-throughput sequencing and SNP genotyping and population genetic theory now enable application of markers for guiding seed zone delineation on the scale envisioned here

Marker informed breeding and ecosystem management has changed profoundly with recent advances in genomic technologies and developing theory in population genetics. High throughput DNA sequencing technologies (2nd and 3rd generation) now allow acquisition of complete transcriptome polymorphism databases at very low prices and with great speed. We propose to develop transcriptome polymorphism databases for eight conifer species, some of which have essentially no available genomic resources of any kind to date. Molecular population genetic theory, derived from coalescent theory, will be used to infer departures from selective neutrality in individual genes of the transcriptome using the information imbedded in DNA sequences. Finally, we will design dense (20,000 SNPs) Illumina Infinium genotyping chips to genotype 2000 trees from each of the eight species (16,000 trees in total). We have considerable experience with all these technologies and theories and have a complete bioinformatics and database infrastructure in place to genotype 16,000 trees for 20,000 SNPs, i.e., at a scale that was unthinkable until very recently.

These species capture a broad spectrum of ecological habitats, represent a broad range of adaptive strategies, and are of importance to the well being of the people of California:


Sugar Pine (P. lambertiana)


Limber Pine (P. flexilis)


Singlelead Pinyon (P. monophylla)

Western White Pine (P. monticola)


Bristlecone Pine (P. asistata)


Douglas-fir (Pseudotsuga menziesii)

Whitebark Pine (P. albicaulis)


Foxtail Pine (P. balfouriana)



Ecosystem Genomics and Forest Health Network Meeting, Davis CA March 2008

Participating Organizations

US Forest Service