Cultures of specimens were maintained in the laboratory at 18 ± 0.5 °C and 37‰ salinity and fed weekly with crushed Sphaeroma sp. (Isopoda). Reproductive biology experiments were carried out on one population for each of the OTUs proposed by Casu and Curini-Galletti (2004) (and unpubl. data): ALo (OTU A), CSo (OTU B), PPx (OTU C), CRo (OTU D), PUo (OTU E), KFo (OTU F). They all showed good survival and remarkable longevity in the laboratory. Two sets of experiments were carried out:
a) Intra-population crossings:
Experiments were carried out by isolating pairs of newborns in 20 ml containers filled with seawater, under the culture conditions described above. The offspring produced weekly by pairs were counted and relocated. Fifteen replicates (i.e., pairs) were set for each of the six populations. To exclude bias due to ageing of specimens, which results in sparse, episodic events of reproduction (unpubl. data), the experiments were carried out for a 20 weeks period. In case one or both members of the pair died before 20 weeks from the onset of the experiment, the replicate was considered as lost, and a new pair was set as replacement.
b) Inter-population crossings:
Experiment was set as above, but the pair of juveniles belonged to different populations. All possible combinations were tested (i.e., ALo × CSo; ALo × PPx; ALo × CRo; ALo × PUo; ALo × KFo; CSo × PPx; CSo × CRo; CSo × PUo; CSo × KFo; PPx × CRo; PPx × PUo; PPx × KFo; CRo × PUo; CRo × KFo; PUo × KFo), with 15 replicates per combination (for details about abbreviation codes see Appendix 1).
ANOVAs were used to analyse differences in reproductive outputs (i.e., the number of offspring produced) among pairs of specimens of different populations in the 20 weeks period. Cochran’s C-test was used to test the assumption of homogeneity of variance. The SNK test was used a posteriori to compare means.