When analyzing specific questions, we were limited to subsamples
of observations on the following data: 94 male–male agonistic
bouts, 71 successful matings and 98 unsuccessful mating attempts;
successful matings were defined by evidence for actual spermatophore
transfer. For example, when examining the reaction of a female
to a mating attempt, we were limited to 73 of the 169
successful and unsuccessful attempts since we needed video recordings
that could be analyzed field-by-field, in which both female and
attempting male were seen continuously from before the beginning
of an attempt to after its conclusion. When a size comparison was
needed, we were limited to cases in which all animals compared
were recorded in the same frame, parallel to each other, and perpendicular
to the camera. For animal length, we used images in which
squids were seen perpendicular to the camera and to an adjacent object
of known size. In many cases we placed a measuring tape near
the egg mass and up into the water column. When examining the
causation for mating attempt failures, we were limited to 51 recordings
in which female, consort male, and attempting male were all
seen continuously from before the onset of the attempt to after it
was finished. Data were derived from behavioral or from focal animal
samplings (sensu Martin and Bateson, 1986). When analyzing
video forwhich individual identitymay be of relevance (for example,
fighting ormating success of a consort),multiple recordings fromthe
same individual (i.e. derived from focal sampling) were omitted.
However, when examining fight outcomes of a challenger, we
could use data collected while following a consort, as the challengers
differed each time.