1. Douglas-fir Seedlings in the Pacific Northwest:
The Genetics of Drought Adaptation
Erda Celer1 Glenn T. Howe1
Oregon State University, 1Department of Forest Ecosystems and Society
 Abstract
 Materials & Methods (cont’d)
 Results & Conclusion (cont’d)
Douglas-fir (Pseudotsuga menziesii) is a widely distributed, ecologically important, and commercially valuable tree species in North America. However, climate change is expected to adversely impact Douglas-fir trees, and assisted migration may become necessary to lessen the effects of climate change. Because drought stress is one of the projected effects of climate change in the western U.S., it is increasingly important to include drought adaptation traits in breeding programs and in reforestation decisions.
I will discuss the design and results of a study that help increase the understanding about the importance of climatic-driven genetic differences for drought adaptation traits in Douglas-fir. The results of this study will provide useful information for understanding drought, enhancing breeding programs, and potentially adjusting forest management to climate change impacts.
Heritability differed widely among traits
High heritability for bud flush
Genetic gains were estimated for two backward selection scenarios
Table 3. Heritabilities and genetic gains of traits measured across both plantations.
Ht14
Ht15
Hinc
Flush
Sprague and Lost Creek
Heritabilities h2
Individual
0.72
0.08
0.64
Family
0.78
0.79
0.34
0.82
Genetic gains (%)
Gain1-12.5% 44 38 25 65
Gain2-2.5% 63 55 36 93
Figure 1. Sprague site (22 reps; 427 families; 6480 Douglas-fir seedlings).
Large height differences in the greenhouse persist in the field
Low correlation between growth in the greenhouse and drought adaptation traits
 Goals
The overarching goal of this study was to understand the effects of drought on the growth and survival of Douglas-fir seedlings. The long term goal of this study is to increase the understanding of the genetic capacity of Douglas-fir trees to tolerate drought stress, obtain useful information to enhance approaches for genetically improving drought adaptation traits, and enhance approaches for appropriately deploying genotypes from breeding programs. Adding to our understanding of the potential effects of climate change will help provide information for practicing effective assisted migration.
Table 4. Correlations among breeding values for Douglas-fir seedling traits measured at Sprague and Lost Creek.
Ht14
Ht15
Htinc
Flush
SFlush
Sprague
0.97
0.06
-0.13
-0.18
0.28
-0.20
0.23
0.45
-0.31
-0.05
-0.17
0.04
0.29
0.34
0.41
0.42
0.05
Lost Creek
Figure 2. Lost Creek site (17 reps; 293 families; 3449 Douglas-fir seedlings).
 Results & Conclusion
The Sprague site is typically hotter and drier than Lost Creek
 Objectives
Table 1. Geographic and climatic characteristics of the Sprague and Lost Creek plantations. Climate variables were derived from ClimateNA (Wang et al. 2012).
 Experimental conclusions
Obtain baseline measurements and climate data to help in the analysis and interpretation of future measurements in the Drought Hardiness Study
Characterize the quantitative genetics of drought adaptation traits
Determine whether drought adaptation traits are associated with the climatic origin of Douglas-fir seedlings
Test Sites
Category
Unit
Abbr.
Sprague
Lost Creek
Geog. ft
ELEV
1067
2920
Climate °C
MAT
9.8
4.6 mm
MAP
875
1677
day
NFFD
275
210
Julian d.
FFP
184
164
PAS 52
612
EXT
37.5 29
Hypothesis 1
In the first growing season, the drought adaptation traits were genetically controlled
Hypothesis 2
Drought adaptation traits and climate variables are correlated
The Sprague site was hotter and drier compared to Lost Creek, and these conditions adversely affected seedling growth, damage, and survival. This suggests that the Sprague site should be good for screening drought adaptation traits
Hypothesis 3
Early flushing was associated with warmer (NFFD) and drier (MSP) climates
Genotypes (families) from areas with warmer and drier climates were more likely to flush early
Hypothesis 4
Low correlation between Ht14 and Mort
There is no relationship between height at the time of planting (Ht14) and mortality either at Sprague or at Lost Creek
 Hypotheses
Drought adaptation traits of Douglas-fir seedlings are partly determined by genetics
Natural selection for drought adaptation traits has been stronger in areas that are warmer and drier
Because of high leaf areas, tall Douglas-fir seedlings are more prone to damage from drought
Early bud flush in Douglas-fir is a genetically controlled drought avoidance strategy
Abbreviations: ELEV=elevation; MAT= mean annual temperature (°C); MAP=mean annual precipitation (mm); NFFD= number of frost-free days; FFP= frost-free period; PAS= proportion of precipitation as snow; EXT= extreme maximum temperature over 30 years.
The trees at the Sprague site grew less, were more damaged, and had greater mortality than the trees at the Lost Creek Site
Table 2. Statistics for traits measured on families of Douglas-fir seedlings in the Sprague and Lost Creek plantations.
 Materials and Methods
Sprague
Lost Creek N
Mean
StDev
Ht14
427
39.97
7.407
293
42.71
6.532
Ht15
49.19
7.837
52.43
7.248
Htinc
9.21
2.263
9.72
1.908
Flush
415
1.69
0.405
292
2.61
0.558
Sflush
0.06
0.091
0.38
0.182 FD
0.53
0.217
0.16
0.134 SD
0.10
0.120
0.04
0.063 LD
0.19
0.190
0.03
0.061
Mort
0.30
0.194
0.12
0.126
 Research implications
This study is designed to impose substantial drought stress by focusing on sites in southern Oregon. The experiment assess about 10,000 Douglas-fir seedlings from 429 families from western Oregon and Washington that were planted at two sites: Lost Creek and Sprague. Measured variables, which I refer to as drought adaptation traits, included height, second flushing, spring bud flush, damage (foliage, stems, and leaders), and survival. Using the growth data and climate variables from ClimateNA, a climate interpolation program, we develop a model to show how genetics and environmental factors interact. Seedlots are mostly from open-pollinated seeds from first-generation parents in orchards, some of them are half-sib families created by pooling full-sib families, and two of them are woods-run seedlots from southern Oregon. We have ≤ 60 trees per family.
There were large differences in height among families in the greenhouse (Heritabilities in the greenhouse are unusually high)
Randomization should be used in the greenhouse
Variation in height after one year in the field largely reflected differences in growth that occurred in the greenhouse
Height increment is more relevant for understanding the genetics of field growth and drought hardiness
Initial measurements can be used as covariates in later analyses
Either Ht14 or Ht15 can be used as an “initial height” with later height measurements to understand drought adaptation in the field
Bud flush was highly heritable
Flush would be amenable to genetic improvement
Flush would be a good trait to use for making seed transfer distances and assisted migration
Abbreviations: Ht14 (cm) is the height of the seedlings in 2014, which corresponds to the growth of the seedlings in the greenhouse; Ht15 is the height of the seedlings in 2015 at the end of the first growing season. Htinc is the difference between Ht14 and Ht15. Flush is the bud flush score in the spring of 2016; SFlush is the presence or absence of second flushing in September 2016; FD is the percentage of dead foliage; SD is the percentage of sunscald damage on the stem; LD_bin is the leader condition (1=damaged, 0=not damaged); Mort is mortality (1=dead, 0=alive).