ECOLOGY OP ISLE ROYALE. 3C5 



the southern centers of preservation, it is evident that the barren ground 

 types must have traversed this region en route to the northern position 

 which they no^A' occupy. But relicts of this type have not been recognized 

 among the mammals, although it seems very probable that some inverte- 

 brates have lingered. Miller ('97, pp. G-8) evidently considers that the 

 exposed north shore of Lake Superior, shows marked Hudson Bay aflSn- 

 ities, but is not able to decide Avhether or not this area is limited to the 

 Lake coast. Of the five mammals which he lists as showing these north- 

 ern affinities, only two, the Caribou and White-footed Mouse, occur on 

 Isle Royale. Miller evidentlj^ did not recognize any barren ground 

 relicts in the fauna, yet its Hudsonian affinities may belong, in part, 

 to this class. The barren ground relicts, when present in the coniferous 

 forest belt, may be expected to occur in open swamps', talus or other 

 open rock areas or habitats, as these conditions will most nearly ap- 

 proach those of the open barren grounds. 



With the amelioration of the glacial climate, the barren ground forms 

 were replaced by an invasion of the stunted tree growth and its asso- 

 ciated fauna. The coniferous forest association, in all probability, in- 

 vaded the north shore region, not only around the western end of Lake 

 Superior but also from the east, where it lingers even today as a domi- 

 nant type upon the higher mountains, thiis preserving a continuous 

 record to the present day; while to the westward this type lias not lin- 

 gered so far to the south because of the absence of favorable mountain 

 habitats. On account of the present great extent of this biotic type in 

 the east, a more rapid northward extension may have taken place there, 

 but the mountainous character of the counti'y, the various water barriers 

 westward to Niagara, and possibly the longer duration of the ice in the 

 northeast may have retarded this advance, so that a relatively more 

 rapid extension took place from Michigan into southwestern Ontario 

 and around the western end of Lake Superior (cf. Taylor, '05, p. 107, 

 map). It therefore seems quite probable that the north shore region 

 was invaded both from southern Ontario and from around the western 

 end of Lake Superior. 



Keturning now to the immediate origin of the Isle Royale mammal 

 fauna, it is quite evident that with the exception of the bats, this fauna 

 reached the island from the north shore of Lake Superior. There is 

 perhaps another possibility, but one which seems highly improbable, 

 and that is, that the island was stocked from the south shore of the 

 Lake at that time during post-Glacial migrations, when it contained 

 a more boreal type of fauna. But when we consider the fact that the 

 Superior basin since Glacial times has had much the same general form 

 as the present lake, it seems probable that lake currents similar to those 

 of the present lake existed, and under such circumstances the north 

 shore fauna, especially to the eastward, would be favored. The ice 

 bridge between the island and the north shore permits direct communi- 

 cation with that shore during the winter. The method of arrival for 

 various mammals must of course remain largely conjectural, but the 

 following methods seem probable; the bats by direct flight; the Caribou, 

 Hare, Lyn^ and Marten probably over the ice ; the Red-backed and White- 

 footed Mouse, Bed Squirrel and perhaps the Weasels by means of drift- 

 wood and lake currents; the aquatic forms, Muskrat, Mink, Beaver, and 

 perhaps Otter, by swimming. 



