A major question in evolutionary biology has long been—would evolution repeat itself, given the chance? The late Stephen J. Gould (1994) famously suggested that if we could replay the evolutionary tape of life on Earth, it would not repeat itself, but play differently. He indicated that “history involves too much chaos,” with too many chance events playing a role for the evolutionary process to be repetitive. Gould stated that “chains of historical events are so intricate, so imbued with random and chaotic elements, so unrepeatable in encompassing such a multitude of unique objects, that standard models of simple prediction and replication do not apply. The contrasting view, espoused by others is that “within certain limits the outcome of evolutionary processes might be rather predictable” (Morris 1998; see also Stern et al., 2009). We have enjoyed using the analogy of the movie “Groundhog Day”, where the star character (Bill Murray) is forced to relive the same series of events day after day-that is, some aspects of evolution may be “hard-wired.”
So, are certain aspects of the polyploidy process actually “hard-wired”? We were able to use the Tragopogon system to address this question. Using multiple data sets involving DNA markers and chromosomal (FISH/GISH) data that were developed across polyploid populations from nature of separate origin, as well as multiple synthetic polyploid lines, we have asked if aspects of polyploid evolution are hardwired or stochastic. The sources of data used consistently reveal that evolution does repeat itself in these independent origins in Tragopogon. For example, concerted evolution, which results in the homogenization of gene sequences to one type, is a common feature of ribosomal RNA genes. In both Tragopogon mirus and T. miscellus we found that concerted evolution is ongoing, but incomplete. It appears that we have caught this process in the act (Kovarik et al. 2005). In contrast to 80-year-old natural polyploids, F1 hybrids have equal contributions of the diploid parents, as do raw (S0) synthetic polyploids, as well as the earliest natural populations of T. mirus and T. miscellus (based on DNA from herbarium specimens). But in all existing natural populations examined (Malinska et al. 2011) concerted evolution has consistently occurred in these new polyploid lines of separate origin and it has repeatedly operated “against” T. dubius, homogenizing those copies in the direction of the other parent.
One-gene-at-a-time surveys (Tate et al. 2006, 2009b; Buggs et al. 2009a, b; Koh et al. 2010) have revealed that some genes are consistently maintained in both copies, whereas others consistently show evidence of homeolog loss or evidence of gene silencing of the same parental copy. Genomic results mirrored the findings based on a one gene at a time approach for 29 genes. In both data sets homeolog losses were frequent and gene loss was rapid and also biased against one parent, T. dubius (Buggs et al., 2012a,b). There were also clear patterns to gene loss in that certain loci were repeatedly missing in independently formed populations. Observed chromosomal changes (aneuploidy, translocations) in the recently formed Tragopogon allotetraploids and synthetic lines are also repeated across origins (Chester et al., 2012, 201; Spoelhof et al., 2017).
Therefore, based on multiple lines of evidence, we have found that certain aspects of polyploid evolution are hard-wired—that evolution is repeated in separate lineages of independent origin as well as synthetic lines within both tetraploid species.
Homeolog loss in T. miscellus across natural populations of independent origin. 61-79% of duplicates show no loss; the same 21-39% show loss; the loss of T. dubius copies very similar to the loss of T. pratensis in this data set.