Method for selecting conifer trees

Abstract

A method for selecting conifer trees is disclosed. This method relates to the early measurement of tree characteristics for selecting trees most likely to exhibit certain traits as they mature. This method can be used for implementing silvicultural treatments, thinning or further breeding.

Description

FIELD OF THE INVENTION

The present invention relates to a method for selecting conifer trees for improved volume, yield and wood quality.

BACKGROUND OF THE INVENTION

The growth attributes and physical characteristics of a conifer tree contribute to the value assessed that tree after a commercial harvest. Such value may be measured by the stem quality, volume of wood and wood quality measures such as strength and stiffness. The earlier these characteristics can be accurately assessed, the more a given site can be planted with trees that provide larger volume and more desirable wood characteristics per acre.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to a method for selecting conifer trees. Conifer trees grow well in many regions, including Douglas-fir in the Pacific Northwest region of the United States and loblolly pine in the Southeast region of the United States. In the Southern hemisphere, radiata pine grows well.

The method of the present invention may be used in selecting trees for breeding based upon characteristics measured in those trees using the methods of the present invention. By using more elite trees for breeding, one is able to plant trees which are more likely to provide higher volumes of wood and more desirable characteristics.

The method of the present invention may also be used for selecting against trees for silvicultural treatments. This may include thinning or removal trees not selected based upon characteristics measured using the methods of the present invention.

The method of the present invention may also be used for a combination of selecting elite trees for breeding and later use of the method for selecting among the resulting trees for silvicultural treatments.

The method of the present invention may also be used to select a group of trees having a range of a certain characteristic to provide diversity in a site.

One factor which may be measured in the present invention is sinuosity. Generally, sinuosity is measured as the amount of stem deflection from vertical plumb in the tree. Increased amounts of sinuosity in a tree causes warped grain that impacts the quality of lumber and percent of yield of lumber and pulp from that tree.

Another factor which may be measured in the present invention is ramicorn branching or cluster knots. A ramicorn branch is highly angled (usually less than 30° to the trunk or stem).

Ramicorn branching may result from second flushing or damage to the terminal bud. Cluster knots or double whorls may result from second flushes during the same growing season. Both ramicorn branching and cluster knots can reduce the yield and quality of lumber from a tree.

EXAMPLE

Douglas-fir trees were planted on three Weyerhaeuser sites, one near Longview, Wash. (LG), one near Twin Harbors, Wash. (TH), one in Vail, Wash. (VL) and one in the Cascade mountain range in Washington (CA).

The trees were obtained by breeding 71 parents that were a combination of first- and second-generation selections from three first-generation provenances: Cascade, Longview, and Twin Harbors. These select parent were then mated with 4 to 6 “tester” parents from first-generation selections to produce a total of 274 full-sib families. The Genetic Controls (GC) (n=40) were open-pollinated first-generation families that served as a basis of comparison between the first- and second-generation trials. The Field Checks (FC) (n=3) were genetically unimproved seed sources or reference populations used to gauge changes in growth and stem quality through genetic selection.

The full-sib families were arranged as single-tree plots in 6 randomized complete replications. Thus, each site included 6 trees per full-sib family. Genetic Controls were located in 6 sub-blocks within each replication on a site. Each Genetic Control family was represented by 4 single-tree non-contiguous plots within each sub-block. Field Checks were planted in three 99-tree sub-blocks.

Height, diameter, terminal and lateral second flushing, and sinuosity were measured annually from ages 4- to 9-years. Double whorls and ramicorn branching were measured annually from ages 5- to 9-years. Spring bud flush was assessed at 2- to 4-years. Weekly observations were made of the site and when about fifty percent (50%) of the trees had flushed, bud flush was noted and trees were recorded as being early or later spring bud flush trees.

More intensive branch measurements were made on a sub-set of trees at each test site.

These measurements included branch number and angle, total knot area, and the percentage of fall growth due to second flushing. This sub-set of trees included 30 unrelated parents from 4 of the 6 replications as well as all the trees in 4 of the 6 Genetic Control sub-blocks and 1 of the 3 Field Check sub-blocks.

Final stem quality was assessed in 9-year-old trees in 4 of the 6 replications and retrospectively measured annual leader length, second flushing, ramicorn branching, and sinuosity. Results were averaged across years.

TABLE 1
Descriptions of traits measured and number of observations
in the 3 generations of test trees pooled across the four test sites.
	Number of trees
Trait		Genetic	Field	Trait
Abbreviation	Full-sib	Control	Check	Description
Tree Improvement measures
H8	6293	898	923	(dm)
				Tree height at 8-years.
D8	6293	898	923	(mm)
				Tree diameter at breast height (1.3 m) at 8-
				years.
V8B	6293	898	923	(dm3)
				Bole volume at 8-years.
BudFlush2-4	6644	947	958	Phenology: 1 = flushed terminal bud. Average
				based on annual assessments at 2- to 4-years
				when 50% of trees on a site begin spring
				budflush (approximately mid-May).
Terminal5-9	5711	791	822	Stem quality: 1 = second flushed terminal bud.
				Average of annual assessments from 5- to 9-
				years.
Lateral6-9	6377	913	938	Stem quality: 1 = second flushed lateral bud.
				Average of annual assessments from 6- to 9-
				years.
Sinu5-9	6347	902	932	(inches)
				Stem quality: Sinuosity or the amount of stem
				deflection from vertical plumb. Average of
				annual assessments from 5- to 9-years.
DWhorl6-9	6377	913	938	Stem quality: 1 = Double whorl or second
				whorl of branches from second flushing.
				Average of annual assessments from 6- to 9-
				years.
Rami6-9	6322	904	934	Stem quality: 1 = one or more ramicorn
				branches from second flushing. Average of
				annual assessments from 6- to 9-years.
Tree Improvement intensive branch measures
BranchNumber	880	294	621	Total number of branches >5 mm diameter in
				whorl and internode immediately above breast
				height.
BranchAngle	880	294	621	Average branch angle in the whorl immediately
				above breast height. Range was 15° to 37°
				above horizontal.
KnotArea	880	294	621	Total cross-sectional area of all branches
				calculated from total number of branches
				multiplied by the average diameter of branches
				in the whorl and internode immediately above
				breast height.
FallGrowth	880	294	621	Percent of annual height increment due to
				second-flushing. Measured on internode
				immediately above breast height.
Silviculture measures
Leader5-9	4248	594	929	(dm)
				Average of annual height increments from 5- to
				9-years. Assessed retrospectively at 9-years.
SilvFlush3-9	4248	593	929	Stem quality: 1 = branch or internode with
				second-flush. Average from 3- to 9-years
				assessed retrospectively at 9-years.
SilvRami3-8	4248	593	929	Stem quality: 1 = presence of one or more
				ramicorn branches. Average from 3- to 8-years
				assessed retrospectively at 9-years.
SilvSinu3-9	4248	592	929	(inches)
				Stem quality: Sinuosity or the amount of stem
				deflection from vertical plumb. Average from
				3- to 9-years assessed retrospectively at 9-
				years.

Eight-year height was used to eliminate outlier trees from the dataset. Outliers were defined as those with heights that were more than three standard deviations below the average for the site.

Individual tree data was pooled across the four test sites and the data was analyzed with the AS Reml statistical package (available from VSN International Ltd. having an address of 5 The Waterhouse, Waterhouse Street, Hemel Hempstead, Herts, UK HP1 1ES) which fits linear mixed models using restricted maximum likelihood. Factors fitted included the fixed effect of site, the random effects of replicate within test site, the additive effect of each tree as estimated from its phenotype, the numerator relationship matrix, and family effects. Models were fitted both with and without the inclusion of the origin term or parent provenance to determine the origin effect on heritability estimates.

The ASReml model was equivalent to:Trait_i.j.k.l.m=mu+test_i(+Origin_m)+(replication:test)_ij+family_k+tree_l+residual_i.j.k.l.m  II

• • Trait_ijkl represents the phenotypic value of an individual tree (tree l) for the trait under analysis;
  • mu is the fitted overall mean for the trait across all four tests;
  • test_i is the effect of the i^th test site (considered to be a fixed effect);
  • Origin_m is the effect of the m^th origin (also fixed);
  • replication:test_ij, the effect of the j^th block replication within each i^th test site;
  • family_k, the non-additive effect specific to the cross that produced full-sib progeny l. A model was utilised that allowed this effect to be fitted to Set 1 (full-sib) progeny but not Set 2 (half-sib) progeny, as it is assumed that half-sib families do not contain a systematic effect for specific combining ability;
  • tree_l, the additive genetic effect of tree l; and
  • residual_ijkl, the residual error associated with tree l, assumed to be independent of the test site on which tree l is growing.

The additive genetic relationship matrix describes all known relationships among trees, parents and other ancestors, and its inverse elements augment the equations representing a tree's genetic effect. The additive genetic variance, dominance variance and individual heritability were estimated.

In an individual tree model, the additive genetic variation (σ²_A) is given directly by the variance among trees, estimated from all phenotypic information, taking into account known genetic relationships and all other effects in the model.

The variance of family deviations, denoted σ²_F, includes all the non-additive genetic variance components, the majority of which is expected to be dominance variance. Dominance genetic variance (denoted as σ²_D) was approximated by four times the variance estimated among full-sib families when additive genetic effects are accounted for—σ²_D=4σ²_FA relatively minor amount of σ²_F will be due to epistatic genetic effects, with this being included in the estimate of σ²_D as calculated here. Note that if there were any non-genetic effects that make members of a full-sib family more alike, this will contribute to the estimate of dominance variance.

Individual heritability (h²) was calculated as the ratio of additive (σ²_A) to phenotypic (σ²_P) variance among trees—h²=σ²_A/σ²_P.The phenotypic variance was calculated as the sum of the additive genetic variance (σ²_A), variance due to families (σ²_F) and the residual variance (σ²)—σ²_P=σ²_A+σ²_F+σ²

Variance due to replications was not included in the estimation of phenotypic variance because replication effects were accounted for when estimating breeding values. Standard errors of all components were estimated directly by the ASReml program.

Genetic correlations were estimated in multi-variate ASReml analyses using an individual tree model:(Trait_m1Trait_m2Trait_mn)_ijl=mu_m.+test_i.m.+(replication:test)_ijm.+tree_.ijllm.+residual_ijlm.  IIIwhere m₁, m₂ . . . m_n denote the n traits in the multivariate analyses and tree l has a pedigree defined by its known parents and their ancestors. It was not feasible to simultaneous estimate all possible covariances using a maximum likelihood procedure, so the data were analysed in blocks of up to five traits. The additive correlations (r_A) between trait1 and trait2 represent the ratio of additive genetic covariance between the traits (cov_A1A2) over the square root of the additive genetic variances of the same traits (σ²_A1 and σ²_A2)—r_A=(cov_A1A2)/√(σ²_A1σ²_A2)Many traits with repeated measures were averaged over the years of observations for the correlation estimates. A correlation involving one trait with the average of another has the same expectation as the correlation between the first trait and any single component of the second, assuming they are genetically the same trait.

TABLE 2
Genetic correlations comparing year of assessment for
spring budflush (e.g., BudFlush4 = assessment at 4-years).
Bold numbers on horizontal are individual heritability estimates
with standard errors.
	Trait	BudFlush2	BudFlush3	BudFlush4
	BudFlush2	0.43 ± 0.06	0.88	0.81
	BudFlush3		0.41 ± 0.06	0.95
	BudFlush4			0.47 ± 0.06

TABLE 3
Efficacy of one- or two-year assessments of lateral flushing and
sinuosity as indirect measures of double whorls, ramicorn branching, and
sinuosity in a 9-year stem. The table shows individual
heritability estimates and genetic correlations between indirect
measures and average stem quality traits and volume.
	Heritability	V8B	DWhorl6-9	Ramicorn6-9	Sinu5-9
Lateral5	0.18 ± 0.03	0.14	0.92	0.91	−0.02
Lateral6	0.21 ± 0.03	0.06	0.98	0.98	0.01
Lateral5&7	0.29 ± 0.04	0.07	0.95	0.91	0.04
SilvFlush3-9	0.37 ± 0.05	0.22	0.93	0.93	0.03
Sinu5	0.21 ± 0.03	0.13	0.09	−0.02	0.95
Sinu6	0.13 ± 0.02	0.12	−0.03	−0.10	0.99
Sinu5&7	0.28 ± 0.04	0.08	0.09	−0.01	0.99

While the different embodiment of the invention has been illustrated and described, it will be appreciated that various changes can be made therein without departing from the spirit and scope of the invention.

Claims

1. A method for selecting a conifer tree for breeding from a group of conifer trees, comprising: (a) noting the existence of second flushed lateral buds for a conifer tree in the group of conifer trees by its fifth year of growth;(b) noting the amount of sinuosity for the conifer tree by its fifth year of growth;(c) measuring the height of the conifer tree by its fifth year of growth; and(d) selecting the conifer tree for breeding if it has (i) less second flushed lateral buds by its fifth year of growth than a reference population of trees;(ii) less sinuosity by its fifth year of growth than the reference population of trees; and(iii) greater height by its fifth year of growth than the reference population of trees.

2. The method of claim 1, further comprising noting the time of spring bud flush of the conifer tree in the breeding group of conifer trees at its second year of growth, determining the time when about 50% of the trees in the breeding group of conifer trees spring bud flushed and selecting a plurality of conifer trees from the breeding group of conifer trees, wherein the selected plurality of conifer trees comprises a plurality of trees with a spring bud flush earlier than the time when about 50% of the trees spring bud flushed and a plurality of trees with a spring bud flush later than the time when about 50% of the trees spring bud flushed.

3. The method of claim 1, further comprising: noting the amount of sinuosity for the tree in the breeding group of conifer trees by its fourth year of growth; andselecting the conifer tree if it hasno second flushed lateral buds by its fifth year of growth, andless sinuosity by its fourth year than the reference population of trees.

4. The method of claim 1, further comprising; noting the existence of second flushed lateral buds for the conifer tree by its seventh year of growth; noting the amount of sinuosity for the conifer tree by its seventh year of growth; andselecting the conifer tree for breeding if the tree has no second flushed lateral buds by its seventh year and if it has less sinuosity by its seventh year than the reference population of trees.

5. The method of claim 4, further comprising: measuring the height of the conifer tree by its seventh year of growth;measuring the volume of the conifer tree by its seventh year of growth; andselecting the conifer tree for breeding if the tree has no second flushed lateral buds by its seventh year, if it has less sinuosity by its seventh year than the reference population of trees, if the height is greater by its seventh year than the reference population of trees, and if it has a greater volume by its seventh year than the reference population of trees.