Significant genetic divergence and low morphological differentiation between Cycas bifida and C. micholitzii
Previous phylogenetic analysis on major clades of the core Cycassection Stangerioides revealed that C. bifid a and C. micholiztii were not sister species and indicated that the dichotomous leaflet morphology appeared to have evolved independently several times in the genus of Cycas (J. Liu et al., 2018). However, the two species were sister species according to the phylogeny of transcriptome data, which may be related to leaf material selection (Y. Liu et al., 2022). It is because they have similar pinnate bifurcation patterns that transcriptome data show that they are closely related. We tend to believe the previous view (J. Liu et al., 2018). Similarly, our results provide strong evidence that the two leaflet dichotomous species had significant genetic divergence based on DNA sequences and microsatellite datasets. Except for nuclear genes AAT and SAMS , two species-specific groups of haplotypes were recognized in the Network analyses (Figure 2). Evidences can also be found from the genetic differentiation coefficient F ST, which is greater than 0.15 and indicates that significant genetic differentiation exists in the two species (Table 3). Both the results of STRUCTURE and PCoA showed that populations were grouped into two clusters corresponding to the two species (Figure 3a, b).
We attribute this low morphological differentiation and high genetic divergence between the two species to two possible hypotheses. The first one, the differentiation of the two species is caused by long-term geographical isolation. The distribution areas of these two species are separated by the RRFZ which can block the gene flow of populations or species on its both sides (Y. Q. Gong et al., 2015; Yang et al., 2017; Y. Zheng et al., 2016). The Indochina Plate moved about 700 km southeast relative to the South China Block along the Red River fault line dued to the collision of the Indian block with Asia (Tang, 2004). The movement of landmasses along the Red River Gulf likely caused many formerly closely related populations of Cycas to separate from one another, stimulating the production of new species through subsequent isolation (Tang, 2004). Cycas bifida and C. micholiztiimay have been one species once or from the same common ancestor, and they are separated species now with high genetic divergence due to the long-term geographical isolation caused by the tectonic movement along the Red River Gulf. The other, their morphological similarity, such as the pinnae bifurcate in same way, is only the result of convergent evolution.
Of the two possible causes, we tend to believe that it is the long-term geographic isolation with the subsequent local adaptation that caused the two species to differentiate. Because in our study, the two species have significant IBD based on the mantel test (Figure 6c) and very little gene flow based on IMa analysis (Figure 5b), and there is a significant barrier between their distribution areas according to the Barrier analysis (Figure 3c), indicating that they diverged by the geographic isolation. According to Hill, species of Cycas that exhibit morphological features similar to those of phylogenetically related species may hybridize freely with each other (Ken D Hill, Stevenson, & Osborne, 2004)(Hill et al., 2004; Hill, 2008). Therefore,C. bifida and C. micholitzii may have gene flow or gene introgression with the Cycas species in the adjacent region, which results in closer phylogenetic relationship with them. Convergent evolution, on the other hand, is unlikely. In many species ofCycas , it is unlikely that only two of them have such a consistent pattern of pinnate branching, or the pinnate branching cannot be used as the taxonomic character for classification of Cycas . The simulated niche similarity results also showed that niches of the two species were not overlapping and were not identity (Figure 4), which further indicated that convergent evolution couldn’t happen between them. In the future, comparative genomic studies on Cycas species will help to reveal why only the two species have the same pinnate bifurcation pattern among so many Cycas species, and that were possibly important in their origin.