230 Gordon A. Riley 



The major diatom flowering occurred in late winter, as previously reported in Long 

 Island Sound (Riley, 1941) and some other littoral waters off the New England coast 

 (BiGELOW, 1926; Fish, 1925). An early flowering in these latitudes requires that the 

 phytoplankton be confined to a shallow stratum, since it is only in such an environ- 

 ment that the effective light intensity in winter exceeds the threshold necessary for 

 active growth. Hence the flowering is likely to be early in shallow water and to begin 

 later in the deeper water, frequently only after the establishment of vernal stability 

 (BiGELOW, LiLLiCK, and Sears, 1940; Riley, 1942). In a particular location the time 

 of the flowering may vary from year to year in accordance with variations in the amount 

 of light available (Atkins, 1928). However, in comparing the winters of 1953 and 

 1954 in Long Island Sound, there were no marked differences in incident radiation, 

 and another explanation must be sought for the fact that the 1954 flowering was about 

 three weeks earlier. 



S. Conover (1955) found that the midwinter diatom population was dominated 

 by two species, Skeletonema costatum and Thalassiosira nordenskioldii, in about equal 

 proportions. In 1953 Skeletonema dominated the flowering, achieving an average 

 concentration of 36 million cells per litre at the stations sampled on March 9, as 

 compared with one million cells per litre of Thalassiosira. In 1954 Thalassiosira 

 nordenskioldii rose to six million cells per litre on February 17, while Skeletonema 

 achieved a peak concentration of nine million cells the following week. Considering 

 the diff'erence in size of the two species, Thalassiosira clearly dominated the early 

 flowering period in 1954. 



Conover conducted a series of experiments in 1954 in wnich natural populations 

 were bottled and treated to a variety of conditions of light, temperature and nii.rient 

 enrichment. Table I shows a part of the data from the experiment of February 10 

 that is pertinent to the present discussion. It is apparent that Thalassiosira has an 



Table I 



Growth coefficients {fractional increase in cell number per day) o/Skeletonema costatum 



and Thalassiosira nordenskioldii in two-day experimental exposures of natural surface 



water phytoplankton populations to normal seasonal temperature and surface light 



intensity, to reduced light, and to increased temperature 



Temperature Growth coefficients 



" C Skeletonema Thalassiosira 



* Average of four experiments at light intensities 

 ranging from 1 to 24% of the surface value. 



optimum growth rate at low temperatures and can tolerate low light intensities, while 

 Skeletonema requires higher temperature and probably better illumination. Later 

 experiments suggested that the critical temperature determining which species will 

 dominate is between 2-4 and 3-7°C. The mean temperature in the Sound was 3-2° in 

 February 1953, and 3-7'' in March, and the corresponding means for 1954 were 

 respectively 1-7 and 3-5°. Conover concluded that the lower temperature in 1954 



