Although Salomonsen (1954) said that species which breed in the same place tend to winter indifferent geographic regions, there is no evidence for this in the five warblers. More precisely, Salomonsen's statement suggests that a certain
amount of competition might be avoided by having allopatric wintering grounds. Probably the most satisfactory way to test this is to determine whether the five warblers' ranges show less winter overlap than a randomly chosen group of five eastern warblers (Western warblers tend to winter in a different region and hence should not be included). To make definitions precise, two species of warbler were said to have a significant overlap in winter ranges if at least half of one species' winter range Is included in the other's. From the winter range data of Bent (1953) the twenty three species of warbler breeding in Maine (Palmer, 1949) show 253 significant overlaps, i.e., an average of 11 per species. Therefore, the probability that a randomly chosen pair of specie of Maine warblers will show significant winterange overlap is 11/23 or .478. Considering the five species of warbler in the present study, CapMay overlaps with myrtle and possibly black throated green; myrife overlaps with Cape May blackthroated green, and blackburnian; black throated green overlaps with Cape May (possibly), myrtle, and blackburnian; blackburn iaoverlaps with myrtle, black-throated green, an baybreasted; and bay-breasted overlaps with blackburnian. There is thus a total of 10 certain overlaps among the 5 species, or 2 per specie There is thus a mean overlap per pair of species 2/5 or .400 which is quite near the expected
0.478 suggesting that the five species overlabs about randomly. It might be argued that the 2 species of Maine warblers themselves show mutual repulsion in the winter ranges and hencare a poor standard of comparison. That this not so can be seen as follows. The 23 specie have a total of 315 significant summer overlaps range, or 13.7 per species; i.e., 13.7/23 or 0.5% is the probability that a randomly chosen pair species will overlap in summer. As discussed above, the probability is 0.478 that a randomly
chosen pair will overlap in winter. Therefore, if winter range is chosen independently of summer, 0.478 x 0.596 = 0.285 would be the expected probability of significant overlap in bot winter and summer. Significant summer an winter overlaps were recorded in 164 cases, givin 7.1 per species, or a probability of 7.1/23 = 0.309
which is even a very slightly higher figure thaexpected, showing a slight tendency for bird which summer together to winter together Therefore, the Maine warblers do not repel on another in over-all winter range and they are therefore suitable for the comparison made earlier It can be concluded that the five species show
about the amount of overlapping of winter rang that would be expected on a random basis.