Toona ciliata is an important timber species but is recognized as an endangered species at level II in China. Its genetic conservation is of increasing concern. Provenance trials and other breeding programs were conducted to develop seed transfer rules and multiplications. Here, we investigated twenty-nine populations sampled across the natural distribution of the T. ciliata complex using mtDNA and nrDNA ITS (ribosomal internal transcribed spacer) markers. Haplotype diversity was h = 0.190 ± 0.202 and nucleotide diversity was π = 0.000383 ± 0.000536 for mtDNA marker. Nucleotide diversity for ITS sequences was 0.00837 ± 0.000783. Haplotypes exhibited phylogeographic structure in spatial distribution. The extent of genetic differentiation was significant (F_st = 0.6994 ± 0.0079 for ITS and 0.8870 ± 0.0077 for mtDNA marker). Isolation by distance (IBD) and by elevation (IBE) occurred among populations. Phylogenetic relationships from mtDNA marker indicated three genetically distinct regions, each without IBD effects. Compared with pollen flow, seed flow was strongly impeded in the western region, but extensive in the central region, and less impeded in the eastern region. Most populations did not exhibit expansion, with only a few populations showing expansion after bottleneck effects. We discussed a strategy of region-based genetic conservation and proposed to conserve multiple populations in the western and eastern regions and a few populations in the central region.

The prediction of genetic gain from artificial selection in a trait is important in plant and animal breeding. Lush’s classical breeder’s equation (BE) is widely used for this purpose, although it is also applied to predicting evolution under natural selection. The current application of high throughput sequencing techniques potentially allows breeders at the individual gene level to capture both additive and non-additive genetic effects. Here, we provide a comprehensive evaluation of predicting genetic gains from the selection at multiple hierarchical levels of population structure (provenances, families within provenances, and individuals within families within provenances). We discuss the processes that could influence the power of prediction under the classical BE, including genetic drift, natural selection, and gene flow. We extend the classical BE to molecular breeding methods for improving the prediction of genetic gains; they include the conventional breeding approach, marker-assistant selection (MAS), genome-wide association study (GWAS), and genomic selection (GS). Lastly, we discuss the genetic gains from the selection using multi-omics traits, including gene expression and epigenetic traits. Our overall synthesis should contribute to a better understanding of predicting genetic gains from the artificial selection under classical and molecular breeding.

Toona ciliata and Toona sinensis belong to the Toona genus of the Meliaceae family and are important timber species in China. T. ciliata is an endangered species at level II due to overcutting and a low rate of natural regeneration. T. sinensis was cultivated as an economic and nutritious tree for more than 2000 years. The two species differ in flower and leaf morphological traits, reproductive systems, and range size of natural distribution. To reveal the potential molecular basis of these divergences, we examined the similarities and differences in their whole genome sequences. Results indicate that T. ciliata had a higher number of expanded gene families than T. sinensis. The whole genome duplication (WGD) occurred before their speciation. The long-terminal repeats (LTRs) insertion was earlier in the T. ciliata genome (3.2985 ± 2.5007 Mya) than in the T. sinensis genome (3.1516 ± 2.2097 Mya). Twenty-five gene families in the T. ciliata genome were detected to be under positive selection compared with background branches of ten different land species. The T. ciliata genome was highly collinear with the T. sinensis genome, but had low collinearity with the genomes of more distant species. These genomic and evolutionary divergences are potentially associated with the differences between T. ciliata and T. sinensis in terms of their reproductive systems and ecological adaptation.

Neolamarckia cadamba (Roxb.) Bosser is a fast-growing deciduous tree species and belongs to the Neolamarckia genus of the Rubiaceae family. This species has great economic and medical values in addition to being an important timber species for multiple industrial purposes. However, few studies have examined the genetic diversity and population structure in the natural distribution of this species in China. Here, we applied both the haploid nrDNA ITS (619 bp for aligned sequences) and mtDNA (2 polymorphic loci) markers to investigate 10 natural populations (239 individuals in total) that covered most of the distribution of the species in China. The results showed that the nucleotide diversity was $\pi$ = 0.1185 ± 0.0242 for the nrDNA ITS markers and $\pi$ = 0.00038 ± 0.00052 for the mtDNA markers. The haplotype diversity for the mtDNA markers was $h$ = 0.1952 ± 0.2532. The population genetic differentiation was small ($F_{st\left(n\right)}$ = 0.0294) for the nrDNA ITS markers but large ($F_{st\left(m\right)}$ = 0.6765) for the mtDNA markers. There were no significant effects of isolation by distance (IBD), by elevation, and by two climatic factors (annual average precipitation and tem perature). A geographic structure among populations ($N_{st}<G_{st}$) was absent. Phylogenetic analysis showed a highly genetic mixture among individuals of the ten populations. Pollen flow was substantially greater than seed flow ($m_p/m_s \gg$ 1.0) and played a dominant role in shaping population genetic structure. The nrDNA ITS sequences were neutral and all local populations did not undergo demographic expansion. The overall results provide fundamental information for the genetic conservation and breeding of this miraculous tree.

Forest genetic conservation is typically species-specific and does not integrate interspecific interaction and community structure. It mainly focuses on the theories of population and quantitative genetics. This approach depicts the intraspecific patterns of population genetic structure derived from genetic markers and the genetic differentiation of adaptive quantitative traits in provenance trials. However, it neglects possible interspecific interaction in natural forests and overlooks natural hybridization or subspeciation. We propose that the genetic diversity of a given species in a forest community is shaped by both intraspecific population and interspecific community evolutionary processes, and expand the traditional forest genetic conservation concept under the community ecology framework. We show that a community-specific phylogeny derived from molecular markers would allow us to explore the genetic mechanisms of a tree species interacting with other resident species. It would also facilitate the exploration of a species’ ecological role in forest community assembly and the taxonomic relationship of the species with other species specific to its resident forest community. Phylogenetic β-diversity would assess the similarities and differences of a tree species across communities regarding ecological function, the strength of selection pressure, and the nature and extent of its interaction with other species. Our forest genetic conservation proposal that integrates intraspecific population and interspecific community genetic variations is suitable for conserving a taxonomic species complex and maintaining its evolutionary potential in natural forests. This provides complementary information to conventional population and quantitative genetics-based conservation strategies.