Analysis of Spine-Raising in the Male Three-Spined Stickleback

Publisher Website
Google Scholar
Abstract
1. Behaviourists have supposed that raising of the dorsal or ventral sets of spines in the reproductively active male three-spined stickleback reflects varying levels of flight and aggression. The object of this research was to test this hypothesis making a particular study of associations of raising of dorsal and ventral spines in approaches to con-specifics with the variables biting, zag-zagging and escape. Associations of 'opening the mouth while approaching' with these variables were also studied. Biting, zig-zagging and escape are behaviour patterns involved respectively in defense of the territory, courting of females, and retreat from the boundary between two sticklebacks' territories. 2. A preliminary experiment, in which the trends of frequencies of spine-raising, biting and zig-zagging were observed in approaches to a male or courting female test-fish held in a glass tube at a distance of 10, 100, or 280 cm away from a solitary male's nest-site at different times of the reproductive cycle, showed that zig-zagging was most frequent at 100 cm, and biting to males at 10 cm, both on days of the cycle with a completed but empty nest, or on the first two days after receiving clutches of eggs from 3 females. On the third, fourth, fifth, and sixth days after receiving eggs, zig-zagging decreased at all distances; biting decreased at 100 and 280 cm, but not at 10 cm to either males or females. This decrease in biting at the greater distances may indicate a decrease in territory size as the time of hatching approaches. The eggs hatched on the fifth or sixth days, the days of the terminal observations. The proportions of approaches with dorsal or ventral spines raised also decreased on the third, fourth, fifth, and sixth days, resulting in a reciprocal increase in approaches with no spines raised, and inferring a causal connection between raising of both sets of spines with either bites or zig-zags or both. Distance effects on spine-raising to males, a decrease in raising dorsal or ventral sets of spines with increasing distance, were generally opposite to those in approaches to females, an increase in raising dorsal when combined with ventral spines with increase in distance, but conclusions were tentative because of small sample size. Because distance appeared to have an effect, it was controlled in later experiments. 3. Associations of biting and zig-zagging with raising spines were first examined by comparing the percentages of approaches with zig-zags or bites in each category of spine-raising. The reverse procedure, categorizing according to bites or zig-zags and comparing the percentages of approaches with dorsal or ventral spines raised between categories, was also performed. Results showed that raising of dorsal spines was positively associated with zig-zagging, and when zig-zags were present in the approach, negatively with biting. Ventral spine-raising showed a positive association with both zig-zagging and biting, as did also raising of dorsal and ventral spines together, and mouth opening. Mouth opening was combined more frequently with raising both dorsal and ventral spines than any other arrangement of raised or lowered spines. Some of these associations could also be observed when the relative frequencies of each combination of spines raised were plotted against the number of bites or zig-zags occurring during the five-minute presentation of test-fish. Analysis by correlations over five minute test-periods appeared to be less sensitive than the former analysis comparing categorized approaches. 4. Because male sticklebacks approach the boundary between their own and a neighbouring male's territory hesitatingly, stimuli which cause escape appeared to be associated with the boundary. One hundred and twenty cm tanks were set up containing two males, each owning a territory in half of the tank with a boundary at the center. Measurement of escape was attempted by comparing the time one male spent fluttering (swimming up and down) when placed, together with his neighbour, in a container at the boundary between their two territories with the time at a point equidistant from the nest but inside his own territory. Fluttering was greater at the boundary, and though attempts by males to return to the nest to fan (ventilation of eggs) may cause some of the fluttering, even when no eggs are present, fanning variables probably do not account for the differences in fluttering at the two positions. Nor can differences in biting frequency do so because fluttering was longer at the boundary even when tests were classed according to bites and compared within classes. These data support the hypothesis that escape is higher at the boundary than inside the territory. 5. Approaches to test-males in glass tubes placed simultaneously one at the boundary and one inside the territory, at the same places in the tank where fluttering was measured, were compared for proportions of approaches made with various combinations of spines raised. Since biting was more frequent to the tube of the test-male inside the territory, and biting affects spine-raising, the comparison was made only in approaches with bites (non-bite approaches were too infrequent to be used) in order to minimize biting effects in the analysis. Ventral spine approaches (with or without dorsal spines) were more frequent at the boundary, and dorsal spines (alone, or with or without ventral spines raised) showed no significant differences at the two positions. This is evidence that escape is one of the variables causing raising of ventral spines. 6. In the 'Discussion' possible spuriousness and directness or indirectness of associations was considered. Spurious associations, arising from definitions, and from numerical restrictions, and some indirect associations were circumvented by the analysis. Since some evidence indicated that assumptions made in the analysis when selecting only bite approaches for examination of effects of escape...

This publication has 0 references indexed in Scilit: