Mussels Mytilus edulis (МЕ) and M. trossulus (МТ) are species that demand
identification using multilocus genotypes. ME and MT co-occur and hybridize along the coasts of
Kola Peninsula. Previous studies have revealed morphological differences between these species
in the White Sea: the majority of MT bear an uninterrupted dark strip of the prismatic layer under
the ligament on the inner side of the shell (Т-morphotype) while the majority of ME lack this
character (E-morphotype). Hybrids are not numerous (about 18% frequency of occurrence in a
mixed sample), and morphologically are similar to that parent species, whose genes dominate in
their genotypes. This morphological character was first introduced for the purpose of
discrimination between MT and M. galloprovincialis in the Russian Pacific Ocean coast.
However, discriminative power of the character has been poorly tested in other geographical
regions. The goal of our study was to find out whether morphological differences between ME
and MT persist in populations along the Barents Sea coast.
ME and MT were sampled in the White Sea (Kandalaksha Bay, 24 samples from
estuarine habitats, salinity up to 25 ppt) and the Barents Sea (Murman coast, 12 samples
from estuarine habitats, salinity as in the Whites Sea; 14 samples from oceanic habitats,
salinity from 25 to 35 ppt). Mussels were genotyped by four allozyme loci. Contribution
of MT and ME genes into individual genotypes was estimated in terms of individual
scores obtained by means of a model-based Bayesian clustering method implemented in
the program Structure 2.3.4. To assess associations among the genetic (Structure scores),
morphological (T, E-morphotypes) and ecological (oceanic vs. estuarine habitats)
variables we used a generalized linear mixed-effect models (GLMMs). Samples were
considered as random factors in the analyses. The R programming language with lme4-
package was used. We found out that, in the low salinity habitats (White Sea and
estuarine Barents Sea samples) differences in morphotype frequencies between МЕ and
МТ were 67% and 63%, respectively. In oceanic habitats, differences were much lower:
21%, on the average, due to the increased frequency of Т-morphotypes among ME. We
hypothesize that recorded morphological differences between ME from estuarine and
oceanic habitats could be due to the ability of a species to adjust the strength of the shell
by active production of the nacreous layer under conditions of high acidity in estuaries.
Unlike ME, MT does not show such plastic responses. Reliability of ME and MT
identification using specified conchological character is satisfactory high in the estuarine,
and low in oceanic habitats across the Kola Peninsula. The probability of correct
identification (when it is conceivable that a mussel with T-morphotype is MT, and a
mussel with E-morphotype is ME) depends on the taxonomic structure of a sample. E.g.,
in pure population of ME, rare mussels with T-morphotypes will be 100% ME. For the
convenience of mussel researches we present the “T-calculator”: statistical functions
relating the frequency of T-morphotype in population to frequency of MT genes and the
accuracy of ME and MT identification using conchological character. The calculator has
been developed for populations from places with salinity below 25 ppt.
The research was conducted in a frame of governmental project АААА-А19-119011690138-0 by MES of Russia and
project 19-74-20024 by RSF.