Various studies on its life history and ecology 

 are in preparation. The following is a sum- 

 mary of work recently performed with live 

 animals in the laboratory. 



Adults 



1. Adult crabs have been maintained in 

 the laboratory under controlled conditions 

 for several months. 



2. Crabs kept isolated in 2-quart containers 

 molt about every 56 days; when kept in groups 

 in circulating sea water they molt about every 

 38 days. 



3. Females bear eggs during winter and 

 spring. The egg-carrying season ended on 

 April 12, 1960, in the laboratory. 



4. Eggs require from 6 to 22 days to 

 ripen and hatch. 



5. Females may produce more than one 

 brood of larvae a breeding season; associa- 

 tion with males is a prerequisite to egg- 

 laying. 



Larvae 



1. The larval stages have been reared 

 from eggs in the laboratory, and some indi- 

 viduals are still alive as juveniles. With 

 streptomycin and no penicillin in the medium, 

 56 out of 100 newly hatched larvae survived 

 to the megalops stage. Ten larval stages 

 have been recognized in these cultures; only 

 five were thought to exist when a previous 

 study, based on plankton samples, was made. 



2. Durations of larval stages have been 

 established. Development from egg to megalops 

 took 56 days in one case and 66 days in 

 another. 



3. The larvae were fed on Artemia nauplii, 

 and may therefore be carnivorous in nature. 



LIGHT, NUTRIENTS, AND BIOTA: 

 STATISTICAL ANALYSIS OF OCEAN DATA 



This section of the progrann is comparable 

 to that on "Statistical analysis of ocean- 

 atmosphere relationships," which was re- 

 ported above, in that it attempts to show 

 predictive quantitative relationships. In this 

 case, the variables being predicted are biotic, 

 relating to links in the food chain of tuna. 

 Ultimately it is hoped to demonstrate rela- 

 tionships predictive for tuna from variables 

 further back along the energy chain, as indi- 

 cated in the introductory sections. At present, 

 for lack of good quantitative data all along 

 the chain for any part of the eastern tropical 



Pacific, such comprehensive relationships can 

 only be indicated in a qualitative way, e.g., 

 the work in the Gulf of Tehuantepec reported 

 in this paper; these studies are valuable, 

 however, in giving leads as to relationships 

 that might later be investigated by statistical 

 methods. 



Relationships in the Energy Chain up to 

 Zooplankton (R. W. Holmes) 



General remarks .- -The purpose of this 

 section of the work is to vmderstand and 

 predict changes in prinnary production (pro- 

 ductivity) and zooplankton standing crop in 

 terms of variables on which they are probably 

 dependent. 



In selecting variables for measurement and 

 analysis the investigator has been guided by 

 his previous experience in the eastern tropical 

 Pacific and by the work of G. Riley (e.g., Riley, 

 1946) at Yale. The large number of oceano- 

 graphic variables measured at noon stations 

 on STOR cruises, and on Expeditions 

 EASTROPIC and (particularly) SCOPE in 

 1955 and 1956, is indicated below sufficiently 

 for the present purpose. Further details are 

 available elsewhere (Holmes et al., 1957; 

 Holmes et al., 1958; Holmes and Blackburn, 

 1960). 



Analysis of these data is still in its pre- 

 liminary stages. A large number of simple 

 two-variable correlation coefficients were 

 first determined as an aid to later work. 

 Multiple linear regression is now being used, 

 and the existence of a suitable program (by 

 R. C. Sprowls and J. Tauchi, University of 

 California, Los Angeles) has enabled effective 

 use to be made of the IBM No. 709 digital com- 

 puter at Western Data Processing Center, 

 University of California, Los Angeles. This 

 program has made it possible to consolidate 

 the data in a form in which they can easily 

 be inspected for relationships. It is realized 

 that this may not be the nnost appropriate 

 method of analysis, because of possible non- 

 linearity and non-normality; however, both 

 zooplankton and chlorophyll a measurements 

 were log-transformed before use. In the re- 

 sults presented below, * and ** indicate coef- 

 ficients, etc. significant at the 5 percent and 

 1 percent level of probability, respectively, 

 and R'- is the fraction of the sum of squares 

 of the dependent variable Y that is attributable 

 to the regression. 



All the measurements were made at noon 

 stations, and n for each regression indicates 

 the number of stations available. It is empha- 

 sized that the values of all variables at each 

 station were obtained nearly siniultaneously, 

 i.e., within 4 hours. Thus the regressions 

 represent attempts to relate contemporaneous 



33 



