One hypothesis seeming v/orthy of testing was that areas to 

 the lee of the island chain might be higher in zooplanktcn abundance 

 than areas to windward because of the possible enrichment from the 

 littoral waters of the islands and from the upwelling which may theoreti- 

 cally occur in the lee of oceanic islands o As the Hawaiian archipelago 

 extends generally in a southeast-northwest direction^ and sir.ce the pre- 

 vailing tradewinds are from the northeast and the major ocean currents 

 from the east and northeast j, a line connecting the various islands of 

 the group as in figvire 11 divides the island waters into windward and 

 leeward areas o In order to examine statistically;' differences between^ 

 and within^ these major areas, they were each subdivided (figo 11) 

 into six subareasj, three leeward and three windward of the islands o 

 Using an analysis of variances of completely randomized design^ we 

 compared the adjusted 2ooplankton volumes obtained on cruises 10^ 12 j 

 and 17 of the Smith for these six subarsaso From the results of the 

 analysis^ summarized in table 8^, we conclude that there were signi- 

 ficant differences (P < O- 35) among cruises ^ but no significant differ- 

 ences (P> O0O5) between windward and leevrard areas or among the six 

 subareas o From an examination of the means (table 8) it is apparent 

 that, on the windward side, subarea 3 produced the lowest mean on all 3 

 cruises; subarea 2 was intermediate in rank in two of the three cruises, 

 and subarea 1 ranked first in two of the three cruises. On the leeward 

 side, subarea 3 was lowest in tv;o of three cruises i, but subareas 1 and 

 2 fail to follow in any particular order. Therefore, while the summary 

 means for both windward and leeward areas show a trend of slightly in- 

 creasing zooplankton volumes from east to west, the individual cruises 

 do not follow this in all instances o 



Although the seven cruises on virhich this report is based 

 were not properly distributed in time to adequately describe seasonal 

 or annual variations in plankton abundances they provide some informa- 

 tion of interest on differences in zooplankton abundance between the 

 summer and fall seasons and between the years 1951 and 1952o It is evi- 

 dent from table 6 that the mean volumes collected in May 1950 ( Smirh 

 cruise 4) were somewhat larger than the volumes collected in August 1950 

 ( Smith cruise 6)0 The time-adjusted meansj 15ol cco/lOOO m,, for cruise 

 4 and 12 o3 cco/lOOO m,^ for cruise 6 are roughly indicative of the de- 

 gree of change o Smith cruise 10 in July 1951 produced an adjusted 

 mean of 25<,6 cCo/lOOO mo^, which is significantly different (P<0a05) 

 from the meanj, 20>,3 cc«/lOOO mo^^ of Smith cruise 12 in October-November 

 1951o Manning cruise 8^ September-October 1951j, which sampled along 

 just one north-south section west of Oahu (figo 2), was intermediate in 

 time and also in zooplankton volume with a mean of 21 o2 cCo/lOOO mo'^o 



In table 9 we have assembled all data obtained during summer 

 and fall of 1951 for the one section, stations 6 to 10 (as numbered on 

 Smith cruise 10), and through an analysis of variance have examined 

 the data for spatial and short-term variations « Miiie the tests of 

 significance indicate that, there were no significant differences among ' 

 stations (P> Oo05) or among visits (P>0»05), the mean square value 

 for visits is quite near the 0.05 level of probabilityo_^ 



30 



