isotherms because they relate to presence or absence of tuna lar- 
vae (Richards 1969). I divided the area into five regions, which 
were occupied during both cruises, to compare the temperature 
regimes. These areas include the region east of the Lesser Antilles, 
the Caribbean Sea (east of long. 66°W, north of lat. 14°N), the 
central Caribbean Sea (east of long. 76°W, south of lat. 18°N, 
north of lat. 14°N), the western Caribbean Sea (west of long. 
76°W, north of lat. 14°N), and the Yucatan Channel. 
Comparison of the five regions is shown in Table 5. Basically, 
winter temperatures (OTP II) are consistently cooler than summer 
temperatures (OTP I). The mixed surface layer lies deeper during 
winter in all regions except in the western Caribbean Sea. To 
facilitate comparisons further, stations from both cruises which 
occurred at the same geographic location were compared for the 
same temperature data (Table 6). These actual values, rather than 
averages, reveal more strikingly the warmer summer values and 
the shallower thermocline. The depths of the 24° and 20°C 
isotherms reveal their variable natures, which are obscured in the 
averages given in Table 5. The area is characterized by eddies 
created by the westerly flow of water past the Lesser Antilles (Ing- 
ham and Mahnken. 1966). Instead of the Caribbean appearing as a 
rather stable environment, there is evidence that it is quite com- 
plex. Ingham (1968) pointed out that the mixed surface layer in 
the Caribbean does not mean that vertical gradients of conser- 
vative properties are zero. He has shown that this is a complicated 
region with the isothermal layer often quite different from the 
isohaline layer. Ingham (1968) suggested that the mechanism 
creating these structures may include advection—besides cooling 
evaporation, precipitation, wind mixing, and convective mix- 
ing—because of the rather high horizontal transport through this 
region. He further added that other mechanisms include wind 
mixing under conditions of surface heating, but evaporation and 
the ‘‘two-diffusivity’’ convection of Stommel and Fedorov 
(1967), caused by differential vertical transport of heat and salts, 
were negligible. Despite these unusual oceanographic features, the 
environment for larval fish is quite thermally stable with 
temperatures above 18°C at 200 m. Where the nets traveled deeper 
than 200 m, only 12 of 109 stations had temperatures < 16°C, and 
only two tows (both during cruise 7239) reached depths with 
temperatures < 10°C. 
Kinds of Fish Larvae Obtained 
The same basic kinds of fish larvae were obtained on both 
cruises with some interesting exceptions (Table 7). As expected, 
the family Myctophidae was represented at every station and was 
the most numerous kind of larvae. The Gonostomatidae was the 
second most abundant family and occurred at all but two stations 
during OTP I and at all stations during OTP II. The following ad- 
ditional 19 families ranked in the top 15 (for either cruise) in oc- 
currence or abundance: Bothidae, Scaridae, Bregmacerotidae, 
Paralepididae, Gobiidae (ranked fourth in all categories), Scom- 
bridae, Labridae, Scopelarchidae, Gempylidae, Carangidae, Ser- 
ranidae, Evermannellidae, Epigonidae, Nomeidae, Synodon- 
tidae, Engraulidae, Congridae, Callionymidae, and Chauliodon- 
tidae. Of these 19 families, 11 were in the top 15 on each cruise in 
both abundance and occurrence: Myctophidae, Gonostomatidae, 
Bothidae, Gobiidae, Scaridae, Labridae, Paralepididae, 
Bregmacerotidae, Gempylidae, Epigonidae, and Scopelarchidae. 
The Scombridae, Caragnidae, and Serranidae were in the top 15 
in three of the four categories, the Nomeidae in two categories, 
and the Callionymidae, Congridae, Synodontidae, Everman- 
nellidae, Chauliodontidae, and Engraulidae appeared once. A 
total of 88 families were recognized in these collections, and 
several additional families were probably present among the 
unidentifiable specimens. Of the 88 families, 26 occurred on only 
one cruise: Alepisauridae, Xenocongridae, Dysommidae, 
Gadidae, Macrouridae, Eutaeniophoridae, Fistularidae, 
Lophotidae, Diretmidae, Caproidae, Sphyraenidae, Mullidae, 
Polynemidae, Sciaenidae, Blennidae, Brotulidae, Ophidiidae, 
Triglidae, Soleidae, Ostracidae, Diodontidae, Gobiesocidae, 
Ogcocephalidae, Ceratiidae, Linophyrnidae, and Gigantactidae. 
This probably reflects, in part, seasonality and the fact that some 
families are very rare. Ahlstrom (1971, 1972) obtained fewer 
families of fish (about 76) than we did, although both collections 
shared most families in common. The following were absent from 
my collections but present in Ahlstrom’s: Giganturidae, 
Scomberesocidae, Trachipteridae, Ammodytidae, Champsodon- 
tidae, Tetragonuridae, Uranoscopidae, Neoscopelidae, and . 
Trachichthyidae. 
A total of 5,569 actual larvae were collected on OTP I and 
3,928 on OTP II. The top 15 families accounted for 69.4% of the 
total number on OTP I and 73.7% of the total number on OTP 
II. Conversely, in Ahlstrom’s (1971, 1972) EASTROPAC studies, 
10 families accounted for over 90% of the larvae. In my work, the 
Myctophidae accounted for 23.4% of the total on OTP I and 
33.3% on OTP II, whereas the Myctophids accounted for 47.2% 
of the larvae on EASTROPAC I and 52.0% on EASTROPAC 
II. Gonostomatidae (includes Sternoptychidae) comprised 14.9% 
of OTP I and 13.7% of OTP II, compared with 29.2% in 
EASTROPAC I and 25.7% in EASTROPAC II. Of the other 
top 10 families of EASTROPAC J and II: the Bathylagidae were 
important, accounting for 5%, but were unimportant in the OTP 
collections; the Melamphaiidae and Idiacanthidae were also im- 
portant in EASTROPAC collections but unimportant in OTP 
collections; the Bregmacerotidae and Paralepididae were impor- 
tant in both regions as were the Nomeidae, Engraulidae, and 
Scombridae; and the Scaridae, Labridae, and Gobiidae were im- 
portant in the OTP collections but not in EASTROPAC. 
In the OTP collections, 10 families accounted for 1.0% or 
more of the total on OTP I, and 12 families comprised 1.0% or 
more on OTP II. Seasonality produced some striking differences 
between OTP I and II. The Scombridae comprised 2.3% of the 
total on OTP I but did not make the top 15 on OTP II. The 
Nomeidae were unimportant on OTP I but made up 1.8% of the 
total on OTP II. Other families ranked among the top 15 in oc- 
currence on either cruise but were absent from the other cruise: 
Chauliodontidae (15th on OTP II); Evermannellidae (10th on 
OTP II); Carangidae (10th on OTP I); and Callionymidae (15th 
on OTP J). 
Two families representative of the vast coral reef habitat of the 
western central Atlantic—Scaridae and Labridae—were also im- 
portant components of the oceanic ichthyoplankton. The 
Scaridae ranked in the top 10 in both occurrence and numbers on 
OTP I and II. The Labridae ranked in the top 10 in both 
categories on OTP I and in the top 15 on OTP II. Gobiid larvae 
ranked fourth in both categories on OTP I and II because their 
larvae are also highly oceanic, despite being benthic in hab‘t as 
adults. Similarly, the benthic family Bothidae ranked in the top 10 
in both categories on both cruises. Bathypelagic fishes of the 
families Paralepididae, Bregmacerotidae, Gempylidae, Scopelar- 
chidae, and Evermannellidae are not surprising major com- 
ponents of the oceanic ichthyoplankton. 
Comparisons can also be made with two other important 
studies of tropical fish larvae. Houde et al. (footnote 4) studied 
the larvae of the eastern Gulf of Mexico on a series of 18 cruises 
