2S7 



tliaii (Idus that (if llie iniddli' lia\-, owinj' to the effect of atmosplieric tenipL-ratures 

 on so large an expanse of relati\el\- shallow water. 



It appears, however, that in the upper bay salinity rather than temperature is 

 the tleterniining factor influencing reproduction and subsequent settlement of larvae. 

 The delayed date of settlement of lar\ae at Crockett and Port Costa in 1922 as com- 

 pared with 1921 seems imdoubtedly due to the prolonged period of low salinities 

 incident upon the heavy rainfall of the winter of 1921-1922, the effects of which were 

 manifested in increased river discharge until well into the sinnmer of 1922. In 1921 

 a mean salinity of 15 parts per 1000 was attained at Crockett on July 6 (see fig. 93) 

 and the salinity ranged above this most of the time subsecpiently ; while in 1922 (see 

 fig. 94) such salinities <li(l not occur until July 29, and jMior to July 6 for five months 

 salinities had been, excejit at a few scattered intervals, below 5 parts per 1000. 



We conclude then that in the lower bay the commencement of the breeding season 

 depends on the temperature factor, while in the upper bay the salinity factor is more 

 important. In this connection reference should again be made to the detailed studies 

 of the conditions at selected points on San Francisco Bay, in the month of July, 1923 

 (figs. 101-105). The data on dissolved gases and hydrogen-ion concentration having 

 already been discussed (p. 265), the discussion here will be limited to salinity and 

 temperature. 



The July temperatures at Dumbarton average 20.6° C. for both surface and 

 bottom. This is more than 2° C. higher than the average at the Oakland Mole, 5° C. 

 higher than at the Ferry building, and 6.5° higher than at Fort Point. The average 

 salinity at Dumbarton during this period is higher than at the Oakland Mole, and 

 nearly the same as at the Ferry Building. It appears therefore that salinity may be 

 ruled out as a factor of importance in regard to the reproduction of Teredo in the 

 middle and lower ba>-, while the probability is emphasized that temperature is the 

 deciding factor. The thought further suggests itself that the relative scarcity of Teredo 

 navalis along the San Francisco waterfront, and its almost complete absence at Fort 

 Point may be a function of the lower temperature prevailing at these localities. 



At Crockett the temperatures are nearly as high as at Dumbarton, averaging 

 1.5° C. less than at the latter locality during the period in question. It is unlikely 

 that this slight difference in temperature accounts in any considerable degree for the 

 delayed breeding season at Crockett, especially in view of the fact that the tempera- 

 tures prevailing at Crockett are consistently higher than those of the middle bay, 

 where during 1922 the breeding season began a month earlier than at Crockett. 



As the salinity increased considerably between the first and last runs at Crockett 

 (fig. 101), the data for each run should be considered separately. As a majority of the 

 surviving adults are at the lower ends of the piles, and settlement of larvae also is 

 heaviest at the lower levels, the bottom salinities are the most important for our 

 purposes. The surface salinities have been plotted as a matter of interest, to show the 

 range of difference in the conditions at the upper and lower levels of the same piling. 



The average bottom salinity during the first run was 15.3, during the second 16.9, 

 and during the third 19.2 parts per 1000, while the minimal bottom salinities were 

 respectively 10.3, 13.1 and 13.8 parts per 1000. Now survival and settlement of larvae 

 may occur in salinities much lower than this, as observed at Antioch in 1920. It 

 appears therefore that the controlling element in the delayed breeding season in the 

 upper bay generally is not inability of the larvae to survive, but failure of the adults 

 to reproduce until favorable conditions of salinity and temperature have prevailed 

 for some length of time. Following the period of winter flood water, which hardly 

 begins to subside before June, the sur\iving adults apparently require a period of 



