Forestry & Forest Sciences · Forest Ecology & Biology · also Tree Improvement, Forest Tree Breeding, Tree Improvement and Seed Technology
The genetics of forest trees and how breeding programmes select, test and deploy better trees: provenance trials, plus trees, seed orchards, progeny tests, clonal forestry and genetic gain.
In the exams: IFoS Forestry Paper I Section B (tree improvement and seed technology); ICAR AICE-JRF 59 Unit 3; ICAR AIEEA-PG Group 09 (forest genetics and tree improvement)
Forest genetics is the study of heredity and variation in forest trees; tree improvement is the application of genetics, together with silviculture, to produce trees that grow faster, have better form and wood, and resist pests, diseases and climatic stress. Trees differ from most crop plants in ways that shape both subjects: they are long-lived and large, take years to reach reproductive age, are mostly outcrossing and highly heterozygous, grow in wild or barely domesticated populations with large genetic diversity, and are planted on sites that cannot be controlled as closely as a farm field. Gains from breeding therefore come slowly, but because they are multiplied over millions of trees and whole rotations, they are among the most cost-effective investments in plantation forestry.
This manual follows the logic of a tree improvement programme. It begins with the variation found in natural populations and the basic genetics that explain it: inheritance, Hardy-Weinberg equilibrium, drift and gene flow, and the mating systems of trees. It then covers the steps of improvement: provenance testing to choose the right seed source; selection of plus trees and seed production areas; seed orchards; progeny tests and field designs; the quantitative genetics of heritability, selection differential and genetic gain; the use of exotics and land races; clonal forestry; hybridisation and polyploidy; and molecular markers and genomic selection. Seed collection, testing, certification and seed zones are covered in tree seed technology, and the ecological basis of variation in forest ecology.
The subject is examined in the UPSC Indian Forest Service Forestry optional (tree improvement and seed technology), in ICAR's AICE-JRF forestry paper (forest genetics and tree improvement unit) and in ICAR's AIEEA-PG. Questions combine recall of concepts and examples with calculations of allele frequencies, heritability and genetic gain. Values such as heritabilities and gains vary widely by species, trait, population and site; the manual gives examples, not universal figures.
The trees in any forest differ from one another in growth, form, phenology, wood properties and resistance to stress. This phenotypic variation has two sources: genetic differences among trees and differences in the environments they grow in, together with interactions between the two. Separating them is the first task of forest genetics, because only the genetic part is passed to offspring and can be improved by selection.
Genetic variation in widespread tree species is organised at several levels: among species, among geographic populations (provenances), among stands within a provenance, among families (trees sharing parents), and among individuals within families. In most widespread, wind-pollinated, outcrossing conifers and many broadleaves, molecular studies show that the great majority of genetic diversity lies within populations, and only a small fraction distinguishes populations; nevertheless, adaptive traits such as the timing of bud flush and set, frost and drought hardiness often differ strongly among populations because natural selection acts on them.
Geographic variation can be continuous or discontinuous. A cline is a gradual change in a trait along an environmental gradient, such as latitude or elevation; for example, populations from colder, higher or more northern sites often stop growing earlier in the season and are more frost-hardy. An ecotype is a population adapted to a particular habitat, with more abrupt differences from neighbouring populations, a concept introduced by Göte Turesson in the 1920s. Edaphic ecotypes adapted to unusual soils, such as serpentine or limestone, are well known. Both patterns reflect a balance between natural selection, which adapts populations to local conditions, and gene flow, which mixes them.
Variation within populations is maintained by the large size of tree populations, extensive gene flow by pollen and seed, outcrossing, long generation times and, in some cases, balancing selection. This high diversity is the raw material of tree improvement and is also what allows populations to adapt to changing environments; conserving it is one of the aims of gene conservation programmes.
Genetic variation ultimately arises by mutation and is reshuffled by recombination during sexual reproduction. Its distribution among populations is shaped by four evolutionary forces: natural selection, which favours genotypes better adapted to their environment; genetic drift, random change in allele frequencies, strongest in small populations; gene flow (migration), which moves alleles between populations and makes them more similar; and the mating system, which determines how alleles combine into genotypes. Human activities, including logging of the best trees (dysgenic selection or high-grading), fragmentation of populations, and the movement of seed, also alter genetic variation in forests.
Traits are controlled by genes located on chromosomes. Each diploid tree carries two copies (alleles) of each gene, one from each parent. Mendel's laws of segregation and independent assortment describe how alleles pass from parents to offspring for single genes with clear effects, such as some disease resistance genes or marker genes. Most traits of economic importance in trees, such as height, diameter, stem straightness and wood density, are quantitative: they vary continuously and are controlled by many genes, each with a small effect, together with the environment. Their inheritance is studied by quantitative genetics, described later.
Conifers have a distinctive genetic feature useful in research: the megagametophyte, the nutritive tissue of the seed, is haploid and genetically identical to the egg, so the maternal contribution to each seed can be analysed directly. Many conifers have very large genomes; Norway spruce and loblolly pine, for example, have genomes several times larger than the human genome.
In a large, randomly mating population with no selection, mutation, migration or drift, allele and genotype frequencies remain constant from generation to generation, and genotype frequencies are given by the Hardy-Weinberg proportions: for a gene with two alleles at frequencies p and q (where p + q = 1), the genotype frequencies are p² for one homozygote, 2pq for heterozygotes and q² for the other homozygote. If 4 per cent of seedlings in a population show a recessive trait (q² = 0.04), then q = 0.2, p = 0.8, and the frequency of heterozygous carriers is 2 × 0.8 × 0.2 = 0.32, or 32 per cent. Departures from these proportions indicate that one or more assumptions are violated: an excess of homozygotes, for example, suggests inbreeding or population substructure. Hardy-Weinberg is the baseline against which forces of change are measured.
Genetic drift is random change in allele frequencies because each generation is a finite sample of the previous one. Its effect depends on the effective population size (Ne), the size of an ideal population that would show the same amount of drift; Ne is usually smaller than the census number because of unequal sex ratios, unequal family sizes and fluctuations in population size. With Nm males and Nf females, Ne = 4NmNf ÷ (Nm + Nf); for 10 and 40 trees contributing as males and females, Ne = 4 × 10 × 40 ÷ 50 = 32. In small populations drift causes loss of alleles and an increase in inbreeding: the inbreeding coefficient rises by about 1 ÷ (2Ne) per generation, so a population with Ne of 10 gains about 5 per cent inbreeding each generation. Small, isolated tree populations, such as remnants in fragmented landscapes or small seed orchards, are therefore at risk of inbreeding depression and loss of diversity.
Gene flow is the movement of alleles between populations by pollen and seed. Trees have some of the highest levels of gene flow among plants: wind-dispersed pollen of conifers and oaks can travel kilometres, and animal pollinators and seed dispersers also move genes far. Gene flow counteracts drift and local differentiation, spreads new adaptive alleles, and connects fragmented populations. It also brings unwanted pollen into seed orchards and production stands, a problem known as pollen contamination, and can carry genes from planted exotics or genetically modified trees into wild populations, a concern in risk assessment.
Forest trees show a range of sexual systems. Monoecious species carry separate male and female flowers or cones on the same tree, as in most conifers (pines, spruces, firs) and many broadleaves (oaks, birches, casuarinas in part). Dioecious species have male and female flowers on separate trees, as in poplars, willows, ginkgo and many tropical species; dioecy enforces outcrossing. Hermaphroditic species have bisexual flowers containing both stamens and pistils, as in teak, eucalypts, dipterocarps and many legumes. In tree improvement the sexual system determines how controlled crosses are made, how orchards are designed (dioecious orchards need male trees) and what the risk of self-pollination is.
Most forest trees are predominantly outcrossing: their seed results mainly from pollen of other trees. Mechanisms that promote outcrossing include dioecy; dichogamy, the separation in time of pollen release and stigma receptivity within a tree (protandry, when pollen is shed first, as in teak and many eucalypts, and protogyny, when stigmas are receptive first); spatial separation of male and female cones within the crown, as in many conifers where female cones are in the upper crown; and self-incompatibility, genetic systems that prevent self-pollen from fertilising ovules, found in many broadleaves. The outcrossing rate, measured with genetic markers, is the proportion of seeds produced by outcrossing; in many conifers it exceeds 90 per cent in natural stands but falls in isolated trees and sparse stands.
Self-fertilisation, when it occurs, usually causes inbreeding depression: selfed offspring of conifers typically have lower seed set, more empty seed, slower growth and higher mortality, because selfing exposes deleterious recessive alleles. This is why seed should not be collected from isolated trees and why seed orchards are designed so that copies of the same clone are not neighbours.
Wind pollination (anemophily) is typical of conifers and many temperate broadleaves such as oaks, birches and alders. Animal pollination prevails among tropical trees: teak is pollinated mainly by insects, especially bees; eucalypts by insects and birds; many dipterocarps by small insects such as thrips; and some trees by bats. The distance pollen travels and the behaviour of pollinators determine gene flow, the effective number of fathers per tree and the risk of inbreeding in sparse stands.
Trees begin flowering only after a juvenile phase that can last from a few years (eucalypts, acacias) to decades (many conifers and slow hardwoods), and many flower irregularly. Tree breeders shorten breeding cycles by inducing early and abundant flowering: grafting mature scions onto rootstocks, girdling or root pruning, applying gibberellins (particularly mixtures of gibberellins A4 and A7 in many conifers), fertilisation and managing water stress. Controlled pollination is carried out by isolating female flowers or strobili in bags before they become receptive, collecting pollen from chosen fathers, and applying it to the isolated flowers.
A provenance is the geographic origin of a seed lot or population. Because populations are adapted to their local environments, seed from different provenances can differ greatly in survival, growth, form, phenology and resistance when grown at the same site. Choosing the wrong provenance has caused some of forestry's most costly failures, such as plantations that grew well for years and were then destroyed by frost, drought or disease to which the provenance was not adapted. Choosing the right provenance, by contrast, often gives gains in growth of tens of per cent at no extra cost, which is why provenance testing is the first step of most tree improvement programmes. Seed zones and transfer rules, which turn provenance knowledge into practice, are covered in tree seed technology.
In a provenance trial, seed collected from many provenances across a species' range is raised in a nursery and planted together in replicated field trials at several sites representing the planting environments. Each provenance is usually represented by seed bulked from many trees. Trials are measured for survival, growth, form, phenology and other traits over many years. Programmes often proceed in stages: range-wide trials with many provenances to identify broad patterns, restricted trials concentrating on the best regions, and proving trials on a larger scale before seed sources are recommended for operational planting.
Internationally coordinated provenance trials, many organised through the International Union of Forest Research Organizations (IUFRO) and through agencies such as the CSIRO Australian Tree Seed Centre and the Danida Forest Seed Centre, have tested important species across countries: Norway spruce and Scots pine in Europe, lodgepole pine across the northern hemisphere, and eucalypts, acacias, pines, teak and Gmelina in the tropics. Results from these trials still guide seed choice and also provide data for studying adaptation to climate.
When provenances rank differently at different sites, there is a genotype-by-environment (G × E) interaction. A provenance that is best on a moist, fertile site may be poor on a dry site, and vice versa. Strong G × E interactions mean that seed sources must be recommended for particular planting zones rather than for a whole country, and that trials must cover the range of planting sites. Statistical analyses, such as stability analysis and regression of provenance performance on site means, identify provenances that are consistently good (stable) and those that are specifically adapted. Climate change adds a new dimension, because the environment of a planting site will itself change over the rotation; provenance trial data are being used to model which seed sources will suit future climates.
A plus tree, or candidate tree, is an individual of outstanding phenotype selected in a natural stand or plantation for use in a breeding programme. Selection is made for the traits the programme aims to improve: growth (height, diameter or volume), stem straightness, freedom from forking, branch angle and size, natural pruning, crown form, health and freedom from pests and diseases, and sometimes wood properties such as density. Trees are selected in stands of known, suitable provenance, and on uniform sites, so that superiority is more likely to be genetic than environmental.
Two main methods are used. In the comparison-tree method, the candidate is compared with a number of the best neighbouring dominant trees (often five) of the same age on the same site; it must exceed them by set margins for growth and form. In the baseline or regression method, used in uneven-aged or mixed stands where comparison trees are lacking, the candidate is compared with a regression of the trait on age or size developed for the stand or region. Scoring systems combine traits into an overall score. Selection intensity is very high: often only one tree in many thousands is selected. A plus tree remains a phenotypic selection until its offspring are tested; trees whose progeny prove superior are called elite trees.
Because a tree's phenotype reflects its site, age, competition and history as well as its genes, phenotypic selection in natural stands achieves only a part of the possible gain, especially for traits with low heritability such as volume growth. Traits with higher heritability, such as stem straightness, branch characteristics and wood density, respond better to phenotypic selection. Selection in even-aged plantations of known origin, where trees have grown under similar conditions, is more effective than selection in irregular natural forest. These limits are why progeny testing is needed to confirm genetic superiority.
A seed production area (SPA) is a good stand, of suitable provenance and above-average phenotypic quality, converted to seed production by removing inferior trees and trees of other species, opening the canopy to encourage crown development and flowering, and managing the stand for seed. Removing inferior trees before they pollinate means that the remaining trees mate mainly with each other, giving modest genetic gains. SPAs are established quickly and cheaply, supply improved seed while seed orchards are being developed, and are important for species where orchards are not yet available. They should be isolated from poor stands of the same species to reduce pollen contamination, and be large enough to maintain genetic diversity. Seed production areas, seed stands and their place in certification are also described in tree seed technology.
A seed orchard is a plantation of selected trees, isolated or managed to reduce pollination from outside sources, and managed for the production of frequent, abundant and easily harvested seed of improved genetic quality. In a clonal seed orchard (CSO), selected plus trees are propagated vegetatively, usually by grafting scions from their crowns onto rootstocks or by rooted cuttings, and the resulting ramets (copies) of each clone are planted in a designed layout. Clonal orchards flower early because the scions come from mature crowns, and they reproduce the exact genotypes selected. In a seedling seed orchard (SSO), seedlings raised from open-pollinated or control-pollinated seed of selected parents are planted, usually at close spacing, and later thinned (rogued) to leave the best individuals of the best families, using the data from the orchard itself, which therefore doubles as a progeny test. Seedling orchards are used for species that are difficult to graft or that flower young, such as many eucalypts and acacias.
Orchard design aims to maximise random mating among clones and minimise self-pollination and mating among relatives. Ramets of the same clone are spaced as far apart as possible; layouts include randomised designs with restrictions preventing adjacent ramets of the same clone, and systematic or permutated neighbourhood designs that ensure each clone has a range of different neighbours. The number of clones is a compromise between gain (fewer, better clones give more gain) and genetic diversity and seed yield (more clones reduce inbreeding risk and broaden the base); orchards commonly contain a few dozen clones or more, depending on species and programme.
Orchards should be on accessible, flat or gently sloping, fertile sites with a climate that favours flowering, and isolated from stands of the same species by a pollen dilution zone, which for wind-pollinated species may need to be several hundred metres or more. Because isolation is rarely complete, pollen contamination from outside stands can be substantial and reduces genetic gain; it is estimated with genetic markers.
Seed orchards are managed intensively: weed control, fertilisation, irrigation where needed, pest and disease control (especially of cone and seed insects), crown management to keep trees low and accessible, flower stimulation, and in some programmes supplemental mass pollination with pollen from selected fathers to raise gains and reduce contamination. Records of each ramet and seed collection are kept so that seed can be traced to the orchard and, where needed, to the mother clone.
Orchards are upgraded over time. A first-generation orchard contains untested plus trees; when progeny test results become available, the poorest clones are removed (rogued), giving a so-called 1.5-generation orchard. Second- and later-generation orchards are established with the best individuals selected within the best families of progeny tests, giving larger gains. Orchards for each generation are part of a breeding strategy that also maintains a broad breeding population for long-term improvement.
A progeny test evaluates the genetic worth of parent trees by measuring the performance of their offspring. It serves three purposes: to estimate the breeding value of parents, so that the poorest can be rogued from seed orchards; to estimate genetic parameters such as heritability and genetic correlations; and to provide a base population from which the next generation of selections is made. In an open-pollinated or half-sib progeny test, seed collected from each mother tree after natural pollination is tested; each family shares a mother but has many fathers. Half-sib tests are simple and cheap but estimate only the general combining ability of mothers, and are affected by pollen contamination and selfing. In a full-sib progeny test, offspring of controlled crosses between known parents are tested, giving more precise information and allowing the separation of general combining ability (GCA), the average performance of a parent's offspring, from specific combining ability (SCA), the extra performance of particular parent combinations.
Mating designs specify which crosses are made. In a polycross, each parent is pollinated with a mixture of pollen from many others, estimating GCA efficiently. In a factorial (or tester) design, a set of females is crossed with a set of males in all combinations. In a diallel, every parent is crossed with every other, giving full information on GCA and SCA but requiring many crosses; partial diallels reduce the work. In nested or hierarchical designs, each male is crossed with a different set of females. The choice balances the information needed against the labour and time of controlled pollination, which is considerable in trees.
Progeny tests are planted in statistically designed field trials. The randomised complete block design, with each family represented once in every block, is the traditional layout; on variable sites with many families, incomplete block designs such as alpha-lattice designs, which use small blocks that do not contain every family, control site variation better. Plots may be single-tree plots, in which each family is represented by one tree per block (efficient for estimating family means with many replicates, but vulnerable to mortality and to competition between families), or multi-tree row or square plots. Trials are replicated at several sites to measure G × E interaction. Border rows reduce edge effects. Measurements are taken at intervals, and early results are related to later ones through age-age correlations, so that selection can be made as early as reliably possible.
Progeny test analysis today often uses mixed linear models that estimate breeding values for every tree, combining information from relatives and from different trials; best linear unbiased prediction (BLUP) is the standard method.
Heritability measures the proportion of phenotypic variation that is due to genetic differences. Broad-sense heritability (H²) is the ratio of total genetic variance (additive plus non-additive, that is dominance and epistasis) to phenotypic variance; it is relevant when whole genotypes are propagated, as in clonal forestry. Narrow-sense heritability (h²) is the ratio of additive genetic variance to phenotypic variance; it determines the response to selection when trees are reproduced sexually, because only additive effects are reliably passed from parents to offspring. Heritability is a property of a particular population, trait, age and environment, not a fixed constant. In forest trees, narrow-sense heritabilities for growth traits are often low to moderate, around 0.1 to 0.3, while traits such as wood density and stem straightness are often higher; values vary greatly between studies.
The selection differential (S) is the difference between the mean of the selected trees and the mean of the population from which they were selected. The response to selection, or genetic gain (R or ΔG), is the expected difference between the mean of the offspring of the selected trees and the population mean, given by the breeder's equation R = h² × S. If the population mean height is 20 m, the selected trees average 23 m (S = 3 m) and h² = 0.3, the expected gain is 0.3 × 3 = 0.9 m, so the offspring are expected to average 20.9 m.
The selection differential can also be written as the selection intensity (i) times the phenotypic standard deviation (σP), where i depends only on the proportion selected; selecting the top 10 per cent of a normally distributed population gives i of about 1.755. Then ΔG = i × h² × σP. With i = 1.755, h² = 0.25 and σP = 2 m, ΔG = 1.755 × 0.25 × 2 = 0.88 m, which on a population mean of 20 m is a gain of about 4.4 per cent.
Several factors modify the simple formula. When only mothers are selected and pollen comes from unselected fathers, as in collecting seed from selected trees in a natural stand, the gain is halved, because only one parent is selected: with h² = 0.3 and S = 3 m, gain becomes 0.5 × 0.3 × 3 = 0.45 m. Pollen contamination in orchards reduces gain in proportion to the share of outside pollen. Gain per unit time, the gain divided by the generation interval, is what matters for comparing strategies: a gain of 0.88 m achieved over a 15-year breeding cycle is about 0.06 m per year. Strategies that shorten the generation interval, such as early selection, flowering induction and genomic selection, can raise gain per year even if gain per generation falls.
Selection for several traits at once uses indices that weight each trait by its economic value and its genetic parameters; correlated responses, such as changes in wood density when selecting for growth, must be monitored, because the correlation between growth and density is often negative. Finally, gain must be balanced against loss of genetic diversity, which is managed by maintaining large breeding populations and limiting relatedness among selections.
An exotic species is one grown outside its natural range. Exotics dominate much of the world's plantation forestry: eucalypts from Australia in Brazil, India, South Africa, China and elsewhere; Pinus radiata from a small native range in California in New Zealand, Chile and Australia; Caribbean, slash and loblolly pines in the tropics and southern hemisphere; Acacia mangium and other Australian acacias in South-East Asia; Gmelina arborea and teak beyond their native ranges; and poplars and Douglas-fir in Europe. Exotics are used because they grow fast, produce wood suited to industry, and often escape the pests and diseases of their native range for a time.
Their use carries risks. Exotics may be poorly adapted to rare climatic extremes; pests and pathogens may follow them or new ones may adapt to them; they may become invasive, spreading from plantations into natural ecosystems, as Prosopis juliflora has in dry India and East Africa, some pines and acacias have in South Africa, and others elsewhere; and large monocultures can reduce biodiversity, water yield and local benefits. Species introduction therefore proceeds through trials and risk assessment.
Species introduction typically passes through stages: species elimination trials, in which many candidate species are planted on small plots to identify those that survive and grow; species testing trials, larger replicated trials of the promising species; and species proving trials on an operational scale before large-scale planting. Each stage includes provenance comparisons within the selected species. Assessment covers growth, form, wood, survival through extreme years, pest and disease susceptibility and invasiveness. Long-term monitoring is essential, because problems may appear only after many years.
A land race is a locally adapted population of an exotic species that has developed over one or more generations of growth and natural selection in its new environment, usually from a narrow initial introduction. Land races may be better adapted to local conditions than new imports from the native range, and are often the starting point of local breeding programmes. Examples include the Pinus radiata land race of New Zealand, developed from a few introductions in the nineteenth century, and Indian eucalypt populations such as the so-called Mysore gum, a hybrid-derived eucalypt land race long planted in South India. Land races can also carry the disadvantages of a narrow genetic base and inbreeding, so programmes often combine them with new introductions from the best native provenances.
Clonal forestry is the planting of selected, tested clones propagated vegetatively, so that every tree of a clone is genetically identical. Clonal propagation captures the whole genetic value of a selected tree, both additive and non-additive effects, so gains are generally larger than from seed of the same selections, and the resulting stands are uniform in growth, form and wood properties. Clonal forestry is most advanced in eucalypts, where clonal plantations in Brazil from the 1970s onwards produced large gains in yield and wood uniformity for pulp, and where Indian programmes have since deployed clones widely on farms and in plantations; it is also used for poplars, willows, Cryptomeria in Japan, Chinese fir in China, acacia hybrids and radiata pine in some programmes.
Clones are produced by rooted cuttings, mini-cuttings (cuttings from intensively managed hedges or mini-hedges of young stock plants), and tissue culture (micropropagation), covered in forest biotechnology and propagation. Many species root easily only from juvenile material, and the ability to root declines as the tree matures; maintaining juvenility in stock plants by hedging, and rejuvenation by repeated grafting or tissue culture, are therefore important techniques.
Clones are selected from progeny tests or hybrid populations and then evaluated in clonal tests, replicated trials at several sites in which each clone is represented by several ramets. Clonal tests estimate the broad-sense genetic value of each clone and its G × E interaction, so that clones can be matched to sites. Testing takes years, and the best clones are then multiplied for deployment. New clones are continuously tested to replace those that decline in vigour or succumb to pests.
The main risk of clonal forestry is the narrow genetic base of planted stands. A pest or pathogen able to attack a widely planted clone can spread rapidly through genetically identical trees, causing large losses; diseases of eucalypt clones and poplar clones have shown this risk in several countries. Other concerns are dependence on few clones, loss of adaptability to changing climate, and social and environmental issues linked to large monocultures.
Deployment strategies manage these risks. Clones can be planted as intimate mixtures, with several clones mixed within each stand, or as a mosaic of monoclonal blocks, each block a single clone, which simplifies management and harvesting while spreading risk across the landscape. Programmes and some national guidelines set minimum numbers of clones to be deployed at a time and limits on the area planted to any one clone; values differ between countries and species, and there is no universally agreed number. Continuous testing and replacement of clones, monitoring for pests and diseases, and maintenance of a broad breeding population from which new clones are drawn are essential parts of a responsible clonal programme.
Hybridisation between species can combine desirable traits of both parents and sometimes produces hybrid vigour (heterosis), first-generation (F1) hybrids that grow faster than either parent. Important tree hybrids include eucalypt hybrids such as Eucalyptus grandis × E. urophylla, widely planted as clones in Brazil, South Africa, China and elsewhere for their combination of growth and disease resistance, and E. tereticornis × E. camaldulensis and related hybrids in India; poplar hybrids such as Populus deltoides × P. nigra (Euramerican poplars, Populus × canadensis) and many others; Dunkeld larch (Larix × marschlinsii, the hybrid of European and Japanese larch), which grows faster than either parent; the acacia hybrid of Acacia mangium × A. auriculiformis, found naturally in Sabah and used in Vietnam and Malaysia; and pine hybrids such as Pinus elliottii × P. caribaea in Queensland.
Hybrids are produced by controlled pollination, or occasionally collected from natural hybrid zones. Because F1 hybrids are usually more uniform and vigorous than later generations, and because F2 and later generations often show segregation and loss of vigour (hybrid breakdown), hybrids are usually deployed as F1 seed from controlled crosses or, more often, as clones, which fix the hybrid genotype. Hybrid breeding requires pure-species breeding populations of both parents, from which parents with good combining ability for hybrids are selected.
Polyploids carry more than two sets of chromosomes. Polyploidy is common in flowering plants but rare in conifers, where coast redwood (Sequoia sempervirens), a hexaploid, is the best-known natural example. In broadleaved trees, triploid forms of aspen (Populus tremula and P. tremuloides) occur naturally and have been noted for vigorous growth and large leaves, and triploid poplars have been bred in China for pulpwood plantations. Polyploids can be induced by treating seeds, seedlings or flower buds with chemicals such as colchicine, which double chromosome numbers, and triploids can be produced by crossing diploids with tetraploids or by using unreduced gametes. Polyploidy may give larger cells, leaves or fibres and, in triploids, sterility, which can reduce invasiveness of exotics. However, many induced polyploids grow poorly, and polyploid breeding has had limited success in trees outside poplars and a few other genera.
Hybrids and their genes can spread into wild populations of the parent species by pollen, a process called introgression, which may threaten the genetic integrity of native populations, for example of native black poplar in Europe exposed to pollen from planted hybrids. Risk assessment of hybrid deployment considers this gene flow, as well as the pest susceptibility of uniform hybrid clones.
Genetic markers are detectable variations in DNA or proteins that reveal genotypes. Isozymes, variants of enzymes separated by electrophoresis, were the first markers widely used in trees from the 1970s and established much of what is known about genetic diversity and mating systems. DNA-based markers followed: random amplified polymorphic DNA (RAPD) and amplified fragment length polymorphism (AFLP), which are dominant markers; microsatellites or simple sequence repeats (SSRs), which are highly variable, codominant and widely used for fingerprinting and parentage; and single nucleotide polymorphisms (SNPs), single-base differences that are abundant throughout the genome and can be genotyped in very large numbers with arrays or sequencing. Chloroplast and mitochondrial markers, inherited through one parent, trace seed and pollen movement and post-glacial migration routes.
Markers have many uses in tree improvement and conservation. They measure genetic diversity and population structure to guide gene conservation. They estimate outcrossing rates and pollen contamination in seed orchards and natural stands. They identify clones (fingerprinting), check the identity of grafts and the pedigree of crosses, and detect errors in labelling, which are surprisingly common in breeding programmes. They allow paternity analysis of open-pollinated progeny, so that full pedigrees can be reconstructed without controlled crosses. They map quantitative trait loci (QTL), regions of the genome associated with variation in traits, though QTL in trees have often explained small parts of the variation and been specific to particular families. Marker-assisted selection based on a few QTL has therefore had limited impact in forest trees.
The genome of black cottonwood (Populus trichocarpa), published in 2006, was the first tree genome sequenced; genomes of Eucalyptus grandis, Norway spruce, loblolly pine and many others have since been published. Conifer genomes are very large, often more than 20 billion base pairs, which made their sequencing difficult.
Genomic selection, proposed by Meuwissen, Hayes and Goddard in 2001 for animal breeding, uses thousands of markers across the whole genome to predict breeding values. A training population of trees with both phenotypes and genotypes is used to build a prediction model; candidates are then genotyped and selected on their predicted genomic breeding values without waiting for field testing. Because it can be applied to seedlings, genomic selection promises to shorten breeding cycles greatly in long-lived trees and so increase gain per year. Studies in eucalypts, pines, spruces and poplars have shown useful prediction accuracies, which depend on the size of the training population, the relatedness between training and candidate trees, trait heritability and the number of markers. Genetic engineering, including gene editing, is also being researched in trees, but its commercial use is limited by regulation in most countries and by forest certification standards that do not allow genetically modified trees.
A student should be able to explain the sources of variation in trees, calculate Hardy-Weinberg frequencies, effective population size and inbreeding, describe mating systems with examples, and outline every stage of a tree improvement programme from provenance trials to clonal deployment. Practise the breeder's equation and gain calculations, and learn the differences between half-sib and full-sib tests, clonal and seedling orchards, and broad- and narrow-sense heritability. Field visits should include plus-tree selection by the comparison-tree method and inspection of an orchard and a progeny trial.
A forester in a tree improvement or plantation unit selects and records plus trees, establishes and manages seed production areas and orchards, makes grafts and controlled pollinations, lays out and measures progeny and clonal trials, supplies improved seed and clones to nurseries, and keeps the records that link every planted tree to its origin. In plantation companies the forester matches clones to sites and monitors clonal stands for pests and diseases.
A senior tree breeder or programme manager designs breeding strategies for species of national or company importance: choosing breeding and production populations, mating and testing designs, generation intervals and deployment methods; balancing gain against diversity; deciding on the use of exotics, hybrids and clones; setting clone-deployment rules; and building cooperatives that share testing among organisations. Such a manager also handles gene conservation and the regulatory aspects of new technologies.
A researcher with long experience studies the genetic architecture of adaptive and economic traits, applies genomic selection and other molecular tools to shorten breeding cycles, analyses provenance and progeny trials to predict responses to climate change, investigates mating systems and gene flow in fragmented forests, and develops strategies for gene conservation and assisted migration. Such work underpins national tree improvement and genetic resource policies.
This manual gives principles and examples; it does not give species-specific heritabilities, genetic gains, orchard designs, clone-deployment numbers or national regulations, which differ widely between programmes and should be taken from current literature and national guidelines. Vegetative propagation and tissue culture techniques are covered in forest-biotechnology-propagation; seed handling, testing, certification and seed zones in tree-seed-technology; gene conservation strategies only briefly. Genetic engineering and gene editing are mentioned only in outline because regulation and evidence are changing quickly. Statistical methods for analysing trials (mixed models, BLUP) are named but not derived.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.