GARRY OAK ECOLOGY AND RESTORATION IN A CHANGING WORLD

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: MCINTIRE-STENNIS
• Reporting Frequency: Annual
• Accession No.: 0217401
• Project Start Date: Oct 1, 2008
• Project End Date: Sep 30, 2009
• Program Code: [(N/A)]- (N/A)

Recipient Organization
UNIVERSITY OF WASHINGTON
4333 BROOKLYN AVE NE
SEATTLE,WA 98195

Performing Department
Ecosystem Sciences

Non Technical Summary
This proposal falls primarily within priority area 1 (foundation areas of knowledge) for McIntire-Stennis funding. To a lesser extent, it also falls within priority areas 2a (multi-state projects), 2b (forest ecosystem services), and 2d (uncertainty and decision making). Much research focuses on describing differences in species distributions, with a recent focus on distribution changes associated with a changing climate; (Shafer et al. 2001; Parmesan and Yohe 2003; Pearson et al. 2004; McKenney et al. 2007). Climate change will greatly affect the management of forest ecosystems, so understanding and predicting these changes is of significant practical concern (Millar et al. 2007). This proposal uses Garry oak (Quercus garryana) as a model species to examine controls on species distribution. Garry oak is the most drought tolerant tree species in western Washington (Van Pelt 2007), allowing it to colonize areas that will not support more water-dependent conifers. Although it benefits from increased water availability, and can even tolerate periodic flooding, it is generally outcompeted on high productivity sites (Stein 1990). As a result, Garry oak is often found in narrow bands between areas with enough water to support conifers and areas with too little water to support trees, such as prairies and shrub steppe (Figure 1). This narrow ecotone suggests that changes in spatial distribution over time should be particularly detectable in Garry oak (Risser 1995).

Animal Health Component

(N/A)

Research Effort Categories

Basic

100%

Applied

(N/A)

Developmental

(N/A)

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
123	0699	1070	100%

Knowledge Area
123 - Management and Sustainability of Forest Resources;

Subject Of Investigation
0699 - Trees, forests, and forest products, general;

Field Of Science
1070 - Ecology;

Keywords

garry

oak

stands

Goals / Objectives
The objectives of this project are to: (1) Examine the age structure of Garry oak stands in central Washington. (2) Quantify the age structure of Garry oak stands throughout its range, and compare observed changes with predicted changes in Garry oak distribution due to climate change.

Project Methods
During summer 2008, we obtained ~ 400 cores from Garry oak trees in 4 stands in central Washington. Stands are located on a riparian bench and an upland slope at each of Swauk Creek valley and the Tieton River valley. We will age these cores to quantify the stand age structures. Cores will be processed using conventional methods (Stokes and Smiley 1968) and scanned and analyzed using WinDendro software ($10,000 new; already owned by my lab). For cores that missed the pith, we will estimate the number of years between the inner ring of the core and the pith using Applequist's (1958) method. The height at which each core was taken was noted during sampling (average was 40 cm). To account for the time required for a Garry oak seedling to grow to this height, we will destructively sample five seedlings (1-1.4 m tall) per stand and age these seedlings at the root collar and in 5 cm height increments (Gutsell and Johnson 2002). Since we will be using a small sample size to establish these site-specific height-growth curves, we will adjust them based on curves developed for Garry oak in western Washington (Kertis 1986). The WinDendro analyses will provide an age and a ring-width chronology for each tree. We will compare age distributions among stands to identify differences in stand history such as recruitment patterns (distinct cohorts vs continuous establishment). We will also integrate the age data with extant spatial data to analyze relationships between age and spatial pattern. The spatial distribution of oaks will be compared to random, even and clumped distributions using a K statistic (Loosmore and Ford 2006); these comparisons will also be made for oaks within discrete age classes. We will use a spline correlogram to test for positive or negative spatial correlation between oaks of different ages (Bjornstad and Falck 2001). We will use the ring-width chronology to identify release events, detected as increases in growth rates in spatially aggregated trees without simultaneous growth increases in other trees in the stand (Winter et al. 2002). Based on our findings concerning stand history and current stand health, we will develop recommendations regarding the spacing, density, and timing of planting during Garry oak restoration activities. We will also identify restoration treatments that are likely to promote tree growth, cavity formation, and use by wildlife species within maturing oak stands.

Progress 10/01/08 to 09/30/09

Outputs
OUTPUTS: We have prepared reports for most land managers summarizing our plot characteristic data, vegetation data, and Garry oak seedling density by size class. The reports also include written and graphical information on how their stand compared with the others. Two undergraduate students received training in the implementation and analysis of research projects, presentation of research results, and report preparation. PARTICIPANTS: Key participants were Laura Blume (graduate student), Conor OMalley (undergraduate student), Scott Batiuk (undergraduate student), and Jonathan Bakker (faculty advisor). This project provided training in dendrochronology and implementation, analysis, and presentation of a research project to two undergraduates (OMalley and Batiuk). TARGET AUDIENCES: Our target audiences are those that are managing lands with existing Garry oak stands. We are interested in improving understanding of how to ensure continued regeneration of Garry oaks to replace older trees as they die, without developing overly dense stands. Our research is particularly useful to those managing the very lands where we did our research, including government agencies, a university, and an indigenous Nation. PROJECT MODIFICATIONS: Although we initially intended to select sites that were similar to each other, controlling variables such as slope, elevation, and vegetative cover, we found that the typical Garry oak stand characteristics varied substantially throughout the state of Washington, and we were unable to fully control for these variables. This left us with about as many predictive variables as experimental units, and little power to test variables independently. We also found our proposed methods to be overly time consuming, and therefore only collected and aged seedlings at a subset of sites so we could assess seedling densities at more sites. With regard to seedling densities, this project served as a pilot study identifying possible trends which may best be further addressed individually or even experimentally on smaller geographic scales.

Impacts
We collected data on Garry oak (Quercus garryana) regeneration in 16 Garry oak stands throughout Washington state. We sampled 2-4 plots per stand. We also aged trees and destructively sampled seedlings at a subset of the stands. This project increased our understanding of the complex interactions among factors affecting regeneration. Unfortunately, the small sample size together with the high variability in seedling densities resulted in few statistically significant findings. Garry oak seedling densities were significantly higher east of the Cascades than west of the Cascades; this broadly agrees with the predicted bioclimatic envelope for future conditions favoring Garry oak. Seedling densities were higher in southern parts of the state. Small seedlings (under 1 m tall) were most abundant at mid elevations, while large individuals (over 1 m tall and under 4 cm dbh) were most abundant at the highest elevations. Slope gradient limited seedling density; plots on slopes over 40% never had over about 15,000 small seedlings per hectare or over about 1,000 large seedlings per hectare. Small seedlings were more likely to be found at high densities with high canopy cover, but large individuals were more dense at low canopy cover. This finding suggests that although Garry oak seedlings can establish more easily under a canopy where there is relatively low herb and shrub cover, in order to persist and grow, they do best without an overtopping tree canopy. Dendrochronological analyses demonstrated that age and dbh were correlated at some sites but not at others. Seedling height and tree radial growth rates are much greater west than east of the Cascades, but regeneration has occurred more recently in stands east of the Cascades. Trees in riparian areas had faster radial growth rates than those in upslope areas. Seedling age and height are correlated, though the strength of this correlation varied among sites.

Publications

• Batiuk, S. 2010. Analysis of the relationship between age and diameter at breast height of Garry oak (Quercus garryana) in Washington state. Environmental Science and Resource Management capstone senior thesis, School of Forest Resources, University of Washington, Seattle, WA. 35 p.
• OMalley, C. 2010. Using stem analysis to evaluate seedling dynamics in Quercus garryana. Unpublished research paper. School of Forest Resources, University of Washington, Seattle, WA. 12 p.