Evolutionary analysis for Phragmites ecotypes based on full-length plastomes

Highlights

• We identified little plastome variation between two sympatric ecotypes of Phragmites australis.

• An original analysis process was explored to estimate divergence time of P. australis.

• Plastome phylogeny can objectively reflect the P. australis divergence.

• Phylogenetic relationship of plastome seems not related to phenotypic differences in P. australis.

• P. australis divergent events all occurred at the time of drastic paleoclimate changes.

Abstract

Phragmites australis, a Poaceae species distributed worldwide, has strong environmental adaptability and easily forms ecotypes. To study the process of ecotypic evolution of P. australis, we used two sympatric ecotypes with obvious phenotypic differences—Swamp Reed (SR) and Desert-dune Reed (DR) from the Hexi corridor of China as materials, and estimated divergence times derived from chloroplast genome sequencing. Their plastomes were very similar, with a sizes of 137,660 bp (SR) and 137,647 bp (DR). The results of multiple comparison methods revealed only 13 differences between them, most of which were in non-coding regions. A Poaceae chronogram was reconstructed from Bayesian inference using five fossil records. We optimized this analysis by exploring the input datasets, the fossil records, and the test model to get an estimated ecotypic divergence time of 73 Kya. Two main branches in the evolution of P. australis were outstanding: the American lineage and the European lineage. SR/DR came from the late expansion of the European lineage and separated during the Last glacial period (11.5–80 kya). The intense environmental changes in Quaternary likely caused genetic diversity accumulation, which became the driving force for the DR formation. Our estimated divergence time approach for Phragmites ecotypes will help to explore the phylogenetic status and environmental adaption process of polyploid taxa within Poaceae.

Introduction

Phragmites australis (Cav.) Trin. ex Steud. (common reed), a perennial grass plant belonging to the Arundinoideae of the gramineae family, is widely distributed on all continents except Antarctica. Common reed (henceforth, reed) can easily live in variety of habitats including fresh, brackish, and sea water wetlands, and even in extremely arid areas. The Hexi corridor of Gansu province is located in the arid hinterland of western China and is characterized by low rainfall and high evaporation, but also by large oases and some stable rivers and lakes. This complex habitat contains abundant genetic resources of P. australis. In the early 1990s, Zhang and Chen (1991) have carried out a study on ecotypic differentiation of P. australis in this special region. Through a careful study of the phenotype, they identified and named four ecotypes of P. australis: swamp reed (SR), light salt meadow reed (LSMR), heavy salt meadow reed (HSMR), and desert-dune reed (DR) (Zhang and Chen, 1991). Subsequent studies showed they have significant differences in leaf structure (Zheng et al., 1999), physiology (Zhu et al., 2003) and genetics (Lin et al., 2007). Among them, DR grows and reproduces on sand dunes with extremely low moisture, and has particularly excellent drought tolerance (Fig. 1). Therefore, since its discovery, this reed ecotype DR has been used to study Phragmites’ adaptability to long-term complex stress. However, so far, due to the lack of reed genome information, the molecular mechanism of this powerful adaptability is still unclear. The acquisition of this adaptive ability can be derived from the accumulation of genetic mutations, or from the contribution of phenotypic plasticity. In order to parse this, it is important to reconstruct the pathway of evolutionary differentiation for DR and to ascertain the forces driving evolution.

The morphological characteristics of Arundinoideae species are easily influenced by the environment, which can cause difficulty in intuitively distinguishing species. This phenotypic plasticity makes interspecific and intraspecific classification of Phragmites more difficult. Although the Phragmites is currently classified as the Molinieae, its morphological characteristics are actually similar to some Arundineae species. In particular, their vegetative organs are mostly identical, especially when the four species of Phragmites (NCBI, Taxonomy) are compared. Phragmites australis is highly plastic in its morphology, development and community structure (Haslam, 1970). Classification of Phragmites is based primarily on inflorescence morphology, but even this characteristic displays complex intraspecific diversity. Therefore, it is necessary to explore the intraspecific classification of Phragmites at the molecular level.

Due to the concerns about ecological invasion, the genotyping of P. australis has received increased attention. Driven by global warming and an increase in human activity, as a powerful invasive species, Phragmites has been rapidly invading to various global wetlands since the Great Navigation Epoch (16–17 century). Because they are so difficult to identify, populations cannot be simply classified as native or invasive, so the identity is often ignored until an ecological threat emerges (Saltonstall et al., 2005). By using the sequencing of two chloroplast non-coding regions, Saltonstall (2002) classified more than 300 Phragmites specimens and fresh samples world-wide, and obtained 27 unique haplotypes. Following the same method, An et al. (2012) also used these two chloroplast sequences to classify the haplotype of Phragmites from northern China. Using longitude 110 °E as the boundary, mixed populations of haplotypes O + M were mainly found in western China, while haplotypes M + P were found in eastern China. They also found six previously undescribed haplotypes. Tanaka et al. (2017) further examined the genetic diversity of Phragmites in southwest China, finding the cultivated haplotype P, but that native haplotypes I, Q and U were dominant in this region. In summary, the genetic polymorphism of P. australis community is very complex, and has obvious regional characteristics from China. However, previous works mainly focused on the genetic background of wild populations of P. australis, ignoring the correlation between morphological, physiological characteristics and genetic classification. Our field investigation found that the haplotype classification method that relies on two randomly selected chloroplast DNA fragments still has obvious limitations for the intraspecific classification of Phragmites (Zhang et al., 2020). And for diversified Phragmites, this haplotype classification based on plastid DNA fragments is also difficult to apply to obtain a complete phylogenetic relationship.

In evolutionary botany, the most reliable way to study the natural process of evolution is to use fossil evidence to estimate the divergence times, for example, a Bayesian evolutionary analysis (Bouckaert et al., 2019). Vicentini et al. (2008) analyze the phylogenetic relationship of 97 gramineae plants using the phyB (nuclear gene) and ndhF (chloroplast gene) sequence as input data. Six fossil records were used in this analysis, but the calculated results differed with the postulated age of the fossils. Hardion et al. (2017) used five chloroplast sequences (about 5% of the total length) to calculate the phylogenetic relationship within the Arundinoideae. Their chronogram was anchored by only two fossil records, and gave age estimates far older than the other existing fossil records for the Poaceae. However, even using complete chloroplast genomes, Teisher et al. (2017) found accurate absolute dating problematic where there is a sparse fossil record and heterogeneous molecular evolutionary rates. So, the key to an accurate chronogram lies in the selection of both the most suitable sequence data and accurately dated fossil evidence.

In fact, reed fossils are common in Pleistocene oryctocoenosis, and related records appear in the original archives of certain fossil collections, but few people describe them professionally, and no one has used them in evolutionary estimates. In a fine study, Voorhies and Thomasson (1979) looked at the fossilized florets of the extinct grass genus Berriochloa communis found in the abdominal cavity of a fossilized rhinoceros, which provided direct evidence for the complete differentiation of Arundinoideae in the Miocene (7–10 Mya). However, the fossil evidence for the earliest reed grasses was not clear. Currently, the oldest recognized remains of Phragmites are early Quaternary (2.5 Mya) fossils from Egypt, confidently identified by preserved internal and external structures of rhizomes with attached stem bases (El-Saadawi et al., 1975). Although imperfect, they are enough to be preliminarily identified phylogenetically as Phragmites.

In this study, we first sequenced and accurately annotated the chloroplast genomes of two reed ecotypes (DR and SR, from the Hexi Corridor, China), and obtained the differences between them. We also reannotated five chloroplast genomes of P. australis from North America in Genebank (NCBI, see Materials and Methods). We found that the structural variation of the chloroplast genome among these was very small, especially in the repetitive region, and the variation was mainly concentrated in the single-copy region. For this reason, we tried to perform phylogenetic analysis based on both the whole genome and on single-copy regions only of P. australis chloroplasts. Finally, we selected high variation genes in the single copy region to make an exon-joint geneset, and reconstructed phylogenetic trees and chronograms for the Poaceae and for Phragmites by bayesian inference based on several fossil records. Most importantly, we estimated the divergence time of the two sympatric ecotypes of Phragmites, DR and SR, which laid an evolutionary foundation for the study of physiological differences. In addition, we have also noticed that there is a strong correlation between the rate of generation of the diversity found between our samples of reed populations and the fluctuation of paleoclimate, which provides a new way to understand the evolutionary pathway for the strong environmental adaptability of Phragmites.