Clone Deployment and Selective Harvest Model for Seed Orchards
Advances in seed orchards and clonal mixtures, addressing the challenges and benefits of managing relatives in breeding populations. Discusses the pros and cons, emphasizing group coancestry as a key factor in achieving net gain. Provides insights into strategies for maximizing breeding value while maintaining genetic diversity.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
E N D
Presentation Transcript
RELATIVES IN SEED ORCHARDS AND RELATIVES IN SEED ORCHARDS AND CLONE MIXTURES CLONE MIXTURES A clone deployment and selective harvest model and algorithms A clone deployment and selective harvest model and algorithms for seed orchards and clonal mixtures. for seed orchards and clonal mixtures. Lindgren, Dag.(Swedish University of Agricultural Sciences, Sweden), Danusevicius, Darius. (Lithuanian Forest Research Institute. Lithuania), H gberg, Karl-Anders. (The Forestry Research Institute of Sweden, Sweden), Weng, Yuhui. (New Brunswick Dept. of Nat. Resources, Canada), Hallingb ck, H.R. (Swedish University of Agricultural Sciences, Sweden) IUFRO conference Seoul IUFRO conference Seoul August 28, 2010 http://daglindgren.upsc.se/iufro10/ http://daglindgren.upsc.se/iufro10/
RELATIVES IN SEED ORCHARDS In advanced generations, avoiding relatives In advanced generations, avoiding relatives becomes problematic. becomes problematic. Relatives may be beneficial. Relatives may be beneficial. Avoidance of relatives becomes costly in gain and Avoidance of relatives becomes costly in gain and requires structuring breeding populations. requires structuring breeding populations. Seedling seed orchards are common and always Seedling seed orchards are common and always contain relatives, so relatedness occurs in seed contain relatives, so relatedness occurs in seed orchards today. orchards today. Natural seeds in natural forests contain Natural seeds in natural forests contain matings relatives, relatives are normal in nature. relatives, relatives are normal in nature. matings of of
PROS AND CONS ALLOWING RELATIVES PROS PROS CONS CONS Inbreeding Inbreeding Lower diversity at same Lower diversity at same clone number clone number Unrelated clones require Unrelated clones require structuring the breeding structuring the breeding population (e.g. many population (e.g. many sublines sublines), which reduces ), which reduces gain. gain. Higher gain Higher gain More options More options More clones can be chosen, More clones can be chosen, thus less thus less selfing selfing and more diversity. diversity. and more
GROUP COANCESTRY IS KEY FACTOR! Group Group coancestry coancestry = Average relatedness = Average relatedness (large full sib family =0.25) (large full sib family =0.25) Note self Note self- -coancestry included coancestry included Effective number Effective number = = Status number Status number = half the inverse of group coancestry inverse of group coancestry Group coancestry = Group coancestry = loss of gene diversity loss of gene diversity since breeding started breeding started = half the since
NET GAIN = W net gain iBV p i i = i clone = (genotype) p deployed = proportion i BV breeding value i = group coancestry = 1 W weight (e.g. )
NET GAIN AT GIVEN GROUP COANCESTRY = i W net gain iBV p i If comparing (maximizing) net gain at If comparing (maximizing) net gain at the same group coancestry (status the same group coancestry (status number) the last term becomes number) the last term becomes constant and the weight not needed! constant and the weight not needed!
WEIGHT AND INBREEDING The weight ( The weight (W W) considers both inbreeding depression and ) considers both inbreeding depression and loss of diversity. loss of diversity. Inbreeding seems generally to be a less important constraint Inbreeding seems generally to be a less important constraint for relatives than gene diversity. for relatives than gene diversity. Self Self- -coancestry coancestry (relatedness of clones with themselves) and (relatedness of clones with themselves) and Cross Cross- -coancestry coancestry may be assigned different weights. may be assigned different weights. Selfing Selfing yields few planted seedlings for conifers and yields few planted seedlings for conifers and wherefore wherefore selfing selfing causes little production loss, while mating causes little production loss, while mating among relatives may give higher production loss. among relatives may give higher production loss. In results presented here these weighting options has not In results presented here these weighting options has not been used, but studying weights indicated that inbreeding been used, but studying weights indicated that inbreeding was a less important constraint on relatives than gene was a less important constraint on relatives than gene diversity. diversity.
INVESTIGATED DEPLOYMENT STRATEGIES Maximizing net gain (optimal selection). By Maximizing net gain (optimal selection). By definition best, but other strategies may be definition best, but other strategies may be good enough and simpler. good enough and simpler. Linear deployment (proportions proportional to Linear deployment (proportions proportional to breeding value) allowing relatives or restricting breeding value) allowing relatives or restricting against relatives. against relatives. Truncation selection allowing relatives or Truncation selection allowing relatives or restricting against relatives. restricting against relatives.
The algorithm is able to optimize Number of families Number of families Number of family members Number of family members Number of copies (ramets) of each family Number of copies (ramets) of each family member member Candidate entries can be combined to return highest net gain (=BV - coancestry)
MAXIMIZING NET GAIN (OPTIMAL SELECTION) Individuals are deployed in the proportions maximizing net gain. feed a computer ( solver in EXCEL) Fam1 Fam2 Fam3 Results Fam4 Fam5
SHORT LISTING With many candidates, calculations may With many candidates, calculations may become difficult, uncertain and non become difficult, uncertain and non transparent. transparent. Short Short- -listing the candidates before the final listing the candidates before the final maximization helps! maximization helps!
APPLICATION: COMPARING STRATEGIES WHEN COMPOSING A CLONAL SEED ORCHARD FROM UNRELATED HALF-SIBS Status number = 12, Swedish Scots pine Swedish Scots pine values Linear deployment Linear deployment restricting against relatives was restricting against relatives was good enough if there are more good enough if there are more than double as many unrelated than double as many unrelated candidates as needed! candidates as needed! Other deployment Other deployment alternatives alternatives were were clearly inferior to inferior to optimal optimal! ! clearly Ordered by rank Optimal proportions Linear unrelated Truncation unrelated Linear related Truncation related
APPLICATION ESTABLISHMENT OF NORWAY SPRUCE SEED ORCHARDS Deployment was performed for two new Norway spruce seed Deployment was performed for two new Norway spruce seed orchards in southern Sweden (replacer of current orchards orchards in southern Sweden (replacer of current orchards Runesten and Runesten and G lltofta G lltofta, Finnvid Prescher responsible). , Finnvid Prescher responsible). The candidates were tested clones from controlled crosses. The candidates were tested clones from controlled crosses. Problem with candidates: Different degree of relatedness, in Problem with candidates: Different degree of relatedness, in particular the best parents are heavily represented in progeny particular the best parents are heavily represented in progeny with high BV! with high BV! Breeding value for different characteristics were weighted to get a Breeding value for different characteristics were weighted to get a value index BV for the selections. value index BV for the selections.
APPLICATION ESTABLISHMENT OF SEED ORCHARDS A shortlist was made with 30 candidates with high BV, but A shortlist was made with 30 candidates with high BV, but not sharing the same parent more than 3 times. The short not sharing the same parent more than 3 times. The short- - listed population status number was 17.5. listed population status number was 17.5. Deployment to seed orchard maximizing gain at status Deployment to seed orchard maximizing gain at status number 13 was made by EXCEL function Solver . number 13 was made by EXCEL function Solver . Both growth and wood density could be considerably Both growth and wood density could be considerably improved improved by the optimization procedure. by the optimization procedure. Cross Cross- -coancestry at optimum was 0.01, thus some relatives. coancestry at optimum was 0.01, thus some relatives. 90% (27) of the short 90% (27) of the short- -listed clones were deployed in varying listed clones were deployed in varying proportions, those selected against would have added much proportions, those selected against would have added much to relatedness. to relatedness. Seems the optimum clonal number become higher if some Seems the optimum clonal number become higher if some relatives are selected. relatives are selected.
APPLICATION SELECTIVE HARVEST IN MATURE SEED ORCHARDS Harvesting only the best clones become a common praxis in Sweden Harvesting only the best clones become a common praxis in Sweden as progeny as progeny- -test results become available and orchards become mature test results become available and orchards become mature and produce surplus seeds. There is also a need for two genetic and produce surplus seeds. There is also a need for two genetic fractions, the best for plant production and the second for direct fractions, the best for plant production and the second for direct seeding. seeding. The number of ramets per clone often varies considerable, which The number of ramets per clone often varies considerable, which increases the demand for suitable tools when genetic thinning is increases the demand for suitable tools when genetic thinning is applied. applied. Suggestions were made for selective harvest procedures in the mature Suggestions were made for selective harvest procedures in the mature Scots pine seed orchards of a Swedish forest company. Scots pine seed orchards of a Swedish forest company.
APPLICATION SELECTIVE HARVEST IN MATURE SEED ORCHARDS An assumption of pollen contamination 50% was made. Each An assumption of pollen contamination 50% was made. Each ramet was assumed to give the same pollen and seed contribution. was assumed to give the same pollen and seed contribution. The number of ramets harvested for each clone can be input. Seed The number of ramets harvested for each clone can be input. Seed production, genetic gain and gene diversity appear as outputs. production, genetic gain and gene diversity appear as outputs. The algorithms made it possible to identify solutions where the clone The algorithms made it possible to identify solutions where the clone representation is heterogeneous and non optimal (due to representation is heterogeneous and non optimal (due to establishment constraints or to calamities). establishment constraints or to calamities). The number of ramets per clone can be optimized maximizing genetic The number of ramets per clone can be optimized maximizing genetic value for required seed need at a given status number in seed orchard value for required seed need at a given status number in seed orchard crop by using the EXCEL function (solver). crop by using the EXCEL function (solver). Optimizing selective harvest in a seed orchard with very different clone Optimizing selective harvest in a seed orchard with very different clone contributions have the potential to increase both the gain and gene contributions have the potential to increase both the gain and gene diversity, diversity, Based on such runs of the relevant worksheet recommendations to the Based on such runs of the relevant worksheet recommendations to the company was given. company was given. ramet
APPLICATION CLONE MIXTURE FROM BLACK SPRUCE FAMILIES To deploy clones to plantations can be made in almost the same way To deploy clones to plantations can be made in almost the same way as for a seed orchard. The penalty for high diversity may be chosen as for a seed orchard. The penalty for high diversity may be chosen lower as there is no inbreeding and less concerns with diversity. lower as there is no inbreeding and less concerns with diversity. This application was planning clone mixtures for New Brunswick, This application was planning clone mixtures for New Brunswick, Canada with tested clones from known families. Figures refers to ten Canada with tested clones from known families. Figures refers to ten year volume. year volume. The data was a black spruce clonal test. The trial includes 17 full The data was a black spruce clonal test. The trial includes 17 full- -sib families, on average 10 clones per family. These families were created families, on average 10 clones per family. These families were created by crossings among 12 trees. by crossings among 12 trees. In New Brunswick where are strict restrictions on clonal mixtures, thus In New Brunswick where are strict restrictions on clonal mixtures, thus status number must be high and not relaxed. Only part of an status number must be high and not relaxed. Only part of an acceptable clonal mixture could be derived from the material analyzed. acceptable clonal mixture could be derived from the material analyzed. sib
Efficiencies of selection methods Shortlist: the top 50 (BV) Shortlist: the top 50 (BV) clones without restriction on clones without restriction on relatedness. relatedness. Trunction 50 Linear 45 Optimization Gain in VOL10 (%) 40 Optimization Optimization was considerable superior. At considerable superior. At Ns=5, Ns=5, Optimization Optimization gave 9% larger gain than the larger gain than the Truncation Truncation and and 6% larger than than Linear Linear. . was 35 gave 9% 30 25 6% larger 20 15 1 2 3 4 5 6 7 8 9 10 Ns For Ns=5 For Ns=5, Truncation , Truncation choose12, while choose12, while Optimization Optimization 15 and 18 18 clones. clones. 15 and Linear Linear
Effects by restrictions on family contribution Effects by restrictions on family contribution Shortlists can limit the contributions from individual families Restriction Restriction Clones Clones Ns=5 Ns=5 Ns=8.6 Ns=8.6 Gain CrossCo Gain CrossCo Gain CrossCo Gain CrossCo 1Clone/Fam 1Clone/Fam 2Clone/Fam 2Clone/Fam 3Clone/Fam 3Clone/Fam No No 17 17 34 34 50 50 50 50 37,7 37,7 37.7 37.7 37.7 37.7 37.7 37.7 0.022 0.022 0.025 0.025 0.027 0.027 0.027 0.027 24.3 24.3 25.7 25.7 26.8 26.8 27.1 27.1 0.028 0.028 0.034 0.034 0.035 0.035 0.037 0.037 For a low status number (Ns) For a low status number (Ns) - - around half of the maximal number per family did not matter for gain, but for a status number near the number per family did not matter for gain, but for a status number near the maximum, it was beneficial to have some families well maximum, it was beneficial to have some families well- -represented. Cross Cross- -Coancestry is relatedness among different clones and was not small (first Coancestry is relatedness among different clones and was not small (first cousin =0.0625), thus it can be beneficial to allow some related clones. cousin =0.0625), thus it can be beneficial to allow some related clones. around half of the maximal - - the shortlisted the shortlisted represented.
