DISPERSAL OF LARVAE 



During the 2 or 3 weeks of free-swiniming life 

 the hirvae of C. virginica are more or less passivelj- 

 carried by currents and are \videl.y distributed in 

 coastal waters. Biologists who have studied the dis- 

 tribution of planktonic bivalve larvae (Thoreon, 1946) 

 agree that their swimming is not strong enough to 

 overcome the water movements which transport 

 them far from the spawning grounds. To a 

 certain e.xtent larvae combat the currents by 

 closing their valves and sinking to the lower level 

 of the water colunm or to the bottom. However, 

 observations of the swTmming habits of artificially 

 raised larvae of C. virgimca kept in tall containers 

 in the laboratory show that most of them remain 

 swimming nearly all the time, and only those 

 that appear to be too weak or are infected by 

 fungi settle to the bottom. 



Various methods are used in oyster research to 

 study the distribution of larvae by taking quanti- 

 tative samples, but none are satisfactory, and the re- 

 sults obtained by the different methods are not 

 comparable. A pump for pumping measured 

 volumes of water from different depths, plankton 

 tow samplers of various designs, plankton 

 traps, and bottle collectors of the type described 

 by Thorson are the devices used in the study of 

 vertical distribution of oyster larvae. The plank- 

 ton tow net is most frequently employed. Larvae 

 may be filtered out through screens, or a pre- 

 served sample of water may be placed in a glass 

 cylinder with the bottom drawn into a funnel with 

 a drain cock. The water may be centrifuged at 

 high speed using the Foerst type electric centrifuge 

 designed primarily for the collection of minute 

 organisms that ordinarily pass through the finest 

 mesh of the plankton net. 



Many observers have found that newly attached 

 young oysters far outnumber the free-swmiming 

 hirvae, particularly of the umbo stage, found in 

 plankton samples (Prj'therch, 1924; Galtsoff, 

 Prytherch, and McMillin, 1930; Loosanoff and 

 Engle, 1940). Similar observations concerning the 

 scarcity of larvae of 0. eduHs were reported by 

 Sparck (1925) for Limfjord waters and by Gaarder 

 (1933) for two Norwegian oyster ponds where the 

 oyster larvae were present only in the deeper and 

 saltier layers of water. Observations on the 

 abundance and distrilnition of oyster larvae made 

 m this country and abroad have been adequately 

 reviewed by Korringa (1941). 



The problem of adequacy of plankton sampling 



LAinAL DEVELOPIIEXT AND METAMORPHOSIS 



in relation to the physical and chemical hydrology 

 of the James River, Va., oyster seed bed area was 

 mvestigated by Pritchard (1952, 1953). His 

 calculations show "that the concentration of late 

 stage larvae in the overlying water sufficient to 

 produce the large observed set needs to be, on the 

 average, only about one larva for 100 liters." 

 Since the basic sampling employed in these 

 studies of distribution of oyster larvae was 100 1., 

 the inadequacy of such a sampling technique is 

 obvious and some better automatic sampling 

 methods should be used to clarify these obscure 

 points of larval behavior. 



In estuaries the vertical distribution of larvae 

 seems to depend on changes in the velocity and 

 dn-ection of tidal cuiTents and the vertical salinity 

 gradients. The oyster larvae have a more or less 

 uniform vertical distribution in rearing tanks 

 (Cole and Knight -Jones, 1939) and in the estuaries 

 and bays wherever water mixing has prevented the 

 formation of vertical gradients of temperature and 

 salinity. 



Several observers have attempted to correlate 

 the distribution of larvae with different stages of 

 tides. Julius Nelson, one of the pioneer students 

 of the biology of the larva of C. virginica m New 

 Jersey waters, believed that the larvae could 

 migrate toward land by rismg at the beginning of 

 flood tide and settling to the bottom before the 

 tm-n to the ebb. By this reaction to tidal changes 

 their dispersion in tidal estuaries is avoided. 

 This idea influenced the research of his son, 

 Thm-low Nelson, and his students, who modified 

 and elaborated the original concept (Nelson, 

 1917, 1921; Nelson and Perkins, 1931; Carriker, 

 1951). 



According to these observations, which were 

 made in New Jersey estuaries, the swarms of 

 larvae are distributed along definite lanes up and 

 down stream from the spawning grounds. If the 

 salinity of water is uniform from bottom to surface, 

 the gi-eatest number of larvae is found at the level 

 of the highest current velocity. In bodies of 

 water with distinct salinity stratification the 

 larvae congregate just above the zone of greatest 

 salinity change. Nelson beUeved that the ad- 

 vanced larval stages drop to the bottom and 

 remain near it during slack water, and that the 

 increased salinity of early flood tide stimulates 

 their swimming upward. This performance re- 

 peated at each change of tide enables the larvae 

 to move upstream by progressive stages. This 



369 



