116 Papers from the Department of Marine Biology. 



moisture in the coarse, porous sand, together with the diy winds, soon 

 kills the drifted young mangrove. This has been demonstrated many 

 times in the phj'^siological experiments of the -v^Titer on Rhizophora and 

 explains the absence of the mangTove association in the Tortugas. 



The plants of the Tortugas, hving as they do in such uniformly xero- 

 phj^ic conditions and on such small areas of land, are naturally confined 

 to a small group called a strand flora, on account of characteristic 

 species which occur on all maritime shores and particularly tropical 

 coasts — such genera as Sccevola, and even the species Ipomcea pes-caprce 

 as foimd on all tropical beaches, as far as the remote islands of the 

 Pacific, according to Guppy^ and Scliimper;- the latter, in his Plant 

 Geography,^ in his classification of tropical Httoral formations calls the 

 open sandy formation a "pes-caprse formation"; Harshberger* also 

 includes a great many species found scattered all over the Tortugas in 

 his treatment of sea-strand formations in South Florida, from which, 

 however, the Florida Keys are excluded. 



Broadly, it may be said that the entire flora of the Tortugas is a 

 strand flora, but to the close observer these plants are easily seen to 

 fall into certain associations or groups of several species. The reason 

 for this gi'ouping of species in a fairly uniform flora the writer believes 

 to be due to two factors : first, the prevailing winds, which frequently 

 carry dense salt spray or mist inshore, and certain of these strand 

 plants are better able to withstand this drenching with a rather strong 

 solution of sodium chloride and magnesium and calcium carbonates, of 

 which latter salts the sea-water contains, in this region, fairly high per- 

 centages;^ second, the aggressiveness of certain species and theabiUty 

 to hold a position occupied by them against invading species — for 

 instance, few plants will advantageously invade an area occupied by 

 Opuntia, even though the sm-face is not entirely covered by its sprawl- 

 ing joints. 



The writer, then, recognizes four fairly marked associations in the 

 general Tortugas flora of strand plants: (1) the Uniola community; 

 (2) the Siirimia community; (3) the Opuntia community; (4) the 

 ChamcE^yc£ community. 



The first association, the Uniola community, is rather heterogeneous 

 and varies slightly on different beaches, depending on the conditions in 

 rough weather and the prevailing winds. The lowest on the beach and 

 close to the high-tide limit is Cakile lunceoluta (Wildenow) O. E. Schulz. 

 The dried stalks of this may be seen in July, and diuing August they 



'Guppy, H. B., Obsen-ations of a Naturalist in the Pacific. 1913. 



-Schimper, A. F. W., Die Indo-Malayische Strandflora. 



'Schimper, A. F. W., Plant Geography upon a Physiological Basis. 1903, p. 387. 



^Harshberger, J. W., The Vegetation of South Florida. Wagner Institute of Science, Phila- 

 delphia, 1914, vol. VII, pt. 3. 



*Mayer, A. G., Annual Report of Director of Department of Marine Biology, Carnegie Inst. 

 Wash., 1910, Year Book No. 9, p. 122, quoting from Report on Analj'sis of Tortugas Sea-Water, 

 by Division of Physical and Chemical Research, U. S. Geol. Survey. 



