Premating barriers in young sympatric snail species

Sympatric coexistence of recently diverged species raises the question of barriers restricting the gene flow between them. Reproductive isolation may be implemented at several levels, and the weakening of some, e.g. premating, barriers may require the strengthening of the others, e.g. postcopulatory ones. We analysed mating patterns and shell size of mates in recently diverged closely related species of the subgenus Littorina Neritrema (Littorinidae, Caenogastropoda) in order to assess the role of premating reproductive barriers between them. We compared mating frequencies observed in the wild with those expected based on relative densities using partial canonical correspondence analysis. We introduced the fidelity index (FI) to estimate the relative accuracy of mating with conspecific females and precopulatory isolation index (IPC) to characterize the strength of premating barriers. The species under study, with the exception of L. arcana, clearly demonstrated preferential mating with conspecifics. According to FI and IPC, L. fabalis and L. compressa appeared reliably isolated from their closest relatives within Neritrema. Individuals of these two species tend to be smaller than those of the others, highlighting the importance of shell size changes in gastropod species divergence. L. arcana males were often found in pairs with L. saxatilis females, and no interspecific size differences were revealed in this sibling species pair. We discuss the lack of discriminative mate choice in the sympatric populations of L. arcana and L. saxatilis, and possible additional mechanisms restricting gene flow between them.


Supplement_3. Average relative densities of Littorina snails grouped by site (year) and part of the intertidal zone.
Snails of different species, sex, and maturity are shown by different colours corresponding to the legend. The categories "obtusata" and "saxatilis" include immature or castrated individuals of the corresponding cryptic species group (as such individuals can be unambiguously identified to species group only). Relative densities are presented in the Supplement_4. N

Supplement_5
. The number of copulations in which Littorina snails of a given species were involved. All the categories of pairs were counted (con-and heterospecific, homo-and heterosexual). Ntotal -total number of pairs includes all pairs where any of the partners belonged to a given species. Nactive, Npassivenumber of pairs as active (or passive) partner includes all pairs (homo-and heterosexual) in which active (or passive) partner belonged to a given species. The category L. saxatilis (1.5 row) includes male L. saxatilis with 1.5 rows of penial glands (presumable hybrids of L. saxatilis and L. arcana). The categories "obtusata" and "saxatilis" include immature or castrated individuals of the corresponding cryptic species group.

Supplement_7. Fidelity index (FI) of female Littorina snails grouped by species, intertidal level and year/site.
FI measures the degree of prevalence of mating of females with conspecific males. It is computed as difference of observed and expected frequencies of heterosexual conspecific copulations with females of a given species, divided by the total number of copulations with females of that species. Values of FI vary from -1 -avoidance, through 0 -random mating, to 1assortative mating. Pairs with L. saxatilis 1.5 row males were excluded from the analysis since it is uncertain which females are conspecific to them.

Supplement_8. Partial Canonical Correspondence Analysis of Littorina mating patterns.
The biplots illustrate the effect of the type of the passive partner (its species and sex, blue regular letters) on the ordination of active partners (red bold letters) after the variation explained by the site and year was removed. A -the upper littoral zone, 149 pairs analysed; B -the lower littoral zone, 157 pairs analysed. Red bold letters indicate active partners (males of A -L. arcana, C -L. compressa, F -L. fabalis, O -L. obtusata, S -L. saxatilis, S1.5 -L. saxatilis males with 1.5 penial rows), regular blue letters -passive partners, only group centroids are shown to avoid clutter (the capital letter encodes species as above, "obt"-"obtusata"-species group, "sax" -"saxatilis"-species group; the lowercase letter encodes sex of a passive partner: m -male, f -female, i -immature or trematode-castrated. On B, males of L. fabalis and females of L. arcana were excluded from the analysis as passive partners). Remoteness from zero reflects non-randomness of mating. Supplement_13. Littorina shell size distribution in population grouped by species and year/site. Snail copulation status (not copulating snails, copulating active males, copulating females, and males copulating with males) is shown with colour. A -upper shore level, B -lower shore level. Violins represent the density of data estimated by the kernel method. Width of the violin at any given level is proportional to relative abundance of snails with corresponding shell height. Dots are means and whiskers represent 95% confidence intervals, obtained via nonparametric bootstrap.
The graphs illustrate the interrelations between sizes of copulating snails and the size distribution in populations. The size of mature snails varied more broadly in non-mating individuals (sampled in square plots) than in copulating pairs. Moreover, usually snails of a bigger size were involved in copulations (the only case where this was not obvious was L. compressa in Kiberg 2017).