3.2 | Recent demographic history
The presence of distinct ROH size clusters represent time-stamps for when population bottlenecks have occurred in the recent past (Bosse et al., 2012; Pemberton et al., 2012). Analysis of the overall distribution of ROH lengths in S. catenatus shows evidence for three size classes: short (0.05−0.2 Mb), medium (0.2−0.7 Mb), and long (0.7−7.9 Mb) (Figures S3 and S4). Assuming a recombination rate of 2.8 cM Mb-1 and a generation time of 3 years (see Materials and Methods), we estimated the following time-frames for the demographic events responsible for each ROH size class: short (268−1071 years before present [ybp]), medium (77−268 ybp), and long (7−77 ybp). The latter two time-frames are consistent with possible anthropogenic impacts on this species (Sovic et al., 2019).
The effects of these events can be assessed by comparing differences in class N ROH between species. As such,N ROH for the three ROH size classes (i.e., short, medium, long) was always greater in S. catenatus relative toS. tergeminus , which is consistent with a greater impact of bottlenecks in S. catenatus during the recent past (Figure 4a). Specifically, S. catenatus genomes had a greater number of short (N ROH = 177, SE = 7), medium (N ROH = 107, SE = 4), and long (N ROH = 65, SE = 2) ROH tracts than S. tergeminus genomes (N ROH [short] = 120, SE = 7; N ROH [medium] = 18, SE = 1;N ROH [long] = 6, SE < 1) (P< 0.05 in all cases; Figure 4a). Fold increases in each class suggest that, in S. catenatus , ROH tract gains have been most substantial lately (i.e., 7−77 ybp), followed by ROH tract gains 77−268 and 268−1071 ybp.
We can also use the relative proportion of each size class to measure the relative impact of each event in a way that accounts for overall differences in inbreeding between the species (Figure 4c). Using this approach, small ROHs associated with the oldest bottleneck made up 93% of all ROH segments in S. tergeminus as compared to 77% inS. catenatus (P < 0.05), whereas the relative proportion of medium (S. catenatus : 17%; S. tergeminus : 5%; P < 0.001) and large (S. catenatus : 5%;S. tergeminus : 1%; P < 0.001) ROHs was significantly greater in S. catenatus (Figure 4c). These proportional increases indicate that more recent demographic events have had a substantially greater impact on S. catenatus .
Finally, S. catenatus populations showed few differences, if any, in class N ROH profiles (Figures 4b and S5a-c). Overall, 14, 33, and 0% of N ROH pairwise comparisons among populations were significant (P < 0.05) for short, medium, and long ROH tracts, respectively (Figures 4b and S5a-c). We obtained similar species- and population-level results using relative proportion metrics for each ROH size class (Figures 4d and S5d-f). Broadly speaking, this suggests similar magnitudes of population declines across different populations spanning the range of the species.