Jennings : A Botanical Survey of Presque Isle. 403 



various kinds, especially during the migrating season.^' When once 

 invasion has been accomplished, the formation of zones will rapidly 

 proceed, mainly through the agency of wind.^^ The submerged for- 

 mations and the formations at the edge of the water will depend to a 

 large extent upon water as an agent of distribution, and in most of 

 these formations vegetative dissemination by means of rootstocks and 

 off-shoots of various kinds is also prominent. 



Many of the plants of the sand-plain and its included habitats have 

 disseminules with special adaptations to dispersal by wind : Populus, 

 Salix, Solida}:;o, Aster, etc. In other species the small seeds are 

 probably drifted about with the dry sand, or more or less of the whole 

 plant may be undermined and uprooted and drift about, scattering 

 seeds along its path. It is also highly probable that disseminules of 

 several of the sand-plain species and species of the adjoining thickets 

 are blown over the level expanses of crusted snow during winter : 

 Alnus, Rhus, Myrica, Finns strobus, etc. In most of these species 

 the disseminules are more or less persistent until long into the winter. 



Deglutition by birds is an important means of distribution of the 

 forest and thicket species, but these species do not thus reach the sand- 

 plain to any considerable extent until the vegetation offers perching 

 facilities. A large number of the species of the forest and thicket 

 formations have fruits eaten by birds, many of the fruits being per- 

 sistent during the winter months. Among such species may be men- 

 tioned Junipei-us, Toxicodendron, Arctostaphylos, Rhus, Celastrns, 

 Smilax, Vitis, and Myrica, while of the less persistent fruits eaten and 

 distributed by birds are such species as Fragaria, Rubus, Frunus, Uni- 

 folium, Vagnera, Vaccinium, etc. Juniperus virginiana almost invari- 

 ably first appears on the sand-plain a itw feet to the leeward of a Cot- 

 tonwood, where the drupe was deposited by a bird whose perching 

 position was probably determined by the strong wind. 



As to the dis.semination of the oaks invasion must be due to animal 

 agencies, as no other explanation seems sufficient to explain the 

 appearance of seedlings at considerable distances from trees old enough 



^'Kerner, A. von Marilaun. Trans, by Oliver, F. W. "The Natural History 

 of Plants." Vol. 2: 867-868. 



** As an instance of probable invasion by means of bird migration may be cited 

 Hypericum drumniondii which has a distribution from " Va. to Ga., 111., Iowa, 

 Kansas, and Tex." (Britton, N. L., "Manual," /. r., p. 628), but now occurs in 

 .\shtabula County, Ohio, and at Presque Isle. The minute seeds of this plant might 

 easily be transported long distances in mud adhering to the feet of migrating birds. 



