A Study On Species Differentiation And Local Adaptation Of Invasive Weeds

Introduced species often have the potential to change relatively fast when encountering new environments. Those species that establish, reproduce and distribute successfully and become invasive in the introduced habitats usually experience adaptive evolution to the invaded land. Hence, invasive species are very suitable systems to study species differentiation and local adaptation. Research on the differentiation and adaptation mechanisms of invasive species would be beneficial to understand the phenomena of biological invasions and may help in management attempts.For this study, two invasive weed species were chosen as materials. We collected seeds from their natural populations of both native and introduced land. The differentiation of progenies from different regions was investigated at phenotypic and genetic levels, using ecological, quantitative genetic and genomic approaches. The quantitative genetic approach was also applied to evaluate whether genetic differentiation is caused by drift or natural selection. Possible adaptive mechanisms of the study species to the invaded regions were discussed. The study was composed by the following three parts.1. Phenotypic differentiation was studied in Ambrosia trifida. Experimental populations of progenies from its native populations in US and invasive population in China were grown in a greenhouse common environment. We compared the maternal seed traits, growth and reproductive traits of the experimental plants as well as their nuclear DNA content and responses to water, nutrition and simulated herbivory stresses. The results show that the plants from invaded region received significantly less infestation and may benefit from enemy release. The experimental plants from invasive region grew taller and slimmer at early stages compared to their native region counterparts, but the shape difference disappeared at later stages. This resource allocation pattern may be advantageous for invasive plants to compete against companion species for sunlight and space at early growth stage. Besides, invasive plants reproduced much earlier than native plants in the experiments. This highly significant difference implies that invasive plants of this species have evolved towards shorter life cycles compared to native plants. In the stress-related experiments, native plants and invasive plants showed similar growth and reproductive responses to drought and poor soil. However, growth and reproduction were reduced more in invasive plants than in native plants, which fact lends support to the EICA hypothesis. Meanwhile, compared to native plants the invasive plants exhibited greater phenotypic plasticity in terms of growth performance to maintain a certain level of reproductive performance under stresses and can maximize reproductive performance under better conditions. We suggest that enhanced phenotypic plasticity may have contributed positively to the invasive abilities of invasive genotypes. In addition, hybrids were detected in our sample of invasive plants but not in native plants, so we suspect hybridization of A. trifida with sister species is more frequent in the invaded region. This might provide richer genetic variation for natural selection to act on in the invaded region.2. Transcriptome reflects all the actively expressed genes in a given genotype. To investigate the genetic differentiation between a native genotype and an invasive genotype of A. trifida, we performed de novo transcriptome sequencing of the two genotypes. Roche/454 Titanium platform was used to sequence the cDNA libraries. The results include 206,343 reads from the native genotype and 238,943 reads from the invasive genotype. Through MIRA and Cap3 assembly we got 3 datasets:10,000 native and invasive genotype shared sequences,7,407 native genotype specific sequences and 9,001 invasive genotype specific sequences, which represents the best description of transcriptome of this species currently available. By BLAST against TAIR transcripts dataset, we recognized 6,859 Unigenes from the shared sequence dataset,3,332 and 4,174 Unigenes from the native genotype specific sequence dataset and invasive specific sequence dataset, respectively. Gene Ontology annotation was performed for each group, and around 80% genes were annotated in each group by biological process, cellular component or molecular function. The information would be used to further study the specific genes related to invasive ability of this species and the underlying molecular mechanisms.3. Whether phenotypic variation has a genetic basis and the cause is drift or selection was investigated in flower-size sexual dimorphism of Silene latifolia. Diallel crossing design was made using progenies from an invasive population in US and two native populations in Europe. By comparing the quantitative genetic parameters calculated from cross offspring phenotypes, we infer that the three populations exhibit highly heritable difference in both calyx width and sexual dimorphism of this trait. QST estimates revealed that 72.6%(males) versus 6.9%(females) of the phenotypic variation arises from differences among populations. The between-range QST/FST ratio was 4.2 for males versus 0.40 for females, suggesting that selection has acted on males but not females, and is responsible for phenotypic differentiation among populations in calyx width and its degree of sexual dimorphism. The result provides evidence for adaptive divergence for the study traits in the species instead of random phenotypic evolution.