only be made when conditions were favor- 

 able, and only seven flights were made. 

 Boat lights were counted by two observers 

 on opposite sides of the plane using hand 

 counters. The flight pattern was arranged 

 so that overlap in counts was minimized. 

 Boats tied up at docks in Key West and 

 Marathon were also counted during the 

 flights. Complete counts of the boats in 

 the fishery each night of the flight were 

 obtained in this manner. 



SHRIMP DISCARDS 



Observations of the amount of shrimp 

 discarded were made from the charter 

 vessels, from fishermen's logbooks, and 

 at heading houses. Little discarding was 

 done by the fleet during the period of 

 observation since virtually all sizes caught 

 could be sold. However, some shrimp are 

 always thrown back; for example, damaged 

 shrimp and shrimp that have recently 

 molted are commonly discarded but are 

 usually not abundant. "No" discarding 

 (meaning no conscious rejection of pink 

 shrimp of any size except dannaged indi- 

 viduals) was done from the charter vessels 

 from about February through August. 

 Before and after this period some of the 

 snnaller sizes were thrown away. When 

 this occurred, unselected samples were 

 taken and the whole shrimp were recov- 

 ered, counted, and measured. In addition, 

 a sample of the heads of the shrinnp kept 

 by the crew were recovered, counted, 

 and measured. This permitted an estimate 

 of the total amount of shrinnp discarded 

 from the total catch. Occasionally a cor- 

 rection had to be made, to allow for part 

 of the catch which was iced down without 

 being headed. This occurred when the 

 catch was so large that time was not taken 

 for heading, or if the shrimp were so 

 small as to make heading tedious. 



Two circumstances reduced the use- 

 fulness of data on discards from the 

 charter vessels, hi the first place the 

 captain of the Captain Mack and later of the 

 Danny Boy had a market for small shrimp 

 to be used for fish bait, a market which 

 was not available to most others. Per- 

 haps, as a consequence of this, he kept 

 smaller shrimp than usual. Secondly, the 

 hauls made by the charter boats were of 

 shorter duration than usual in commer- 

 cial practice, resulting in smaller catches, 

 which may have affected discarding prac- 



tices. Hence the data on discards may 

 provide only estimated maximum and 

 mininnunn discard sizes. 



TREATMENT OF DATA 



Size Distribution of Shrimp 



Female pink shrimp are larger than 

 the males and a sample containing both 

 sexes produces a bimodal size distribu- 

 tion. The sex ratio in the total samples 

 approximates 1:1, but there is consider- 

 able variation in the sex ratio among 

 shrimp taken in individual samples. Con- 

 sideration was therefore given to the use 

 of only the females in calculating mean 

 sizes of shrimp found on the fishing 

 grounds. This would, of course, describe 

 only half the population, but if the sizes of 

 both sexes have a linear relation, descrip- 

 tion of the distribution of one sex will 

 provide an approximation of the size dis- 

 tribution of the other sex. The average 

 sizes of males and females of all samples 

 of 30 or nnore shrimp of each sex are 

 plotted in figure 3. All sizes and all depths 

 sampled are included. The figure shows 

 that there is a linear relation between the 

 sizes of the two sexes. Hence the average 

 size of females was selected to illustrate 

 the size distribution of shrimp on the 

 fishing grounds. For comparison four 

 diagrams were constructed using male 

 shrimp only. 



Sampling Variation from Single and 

 Double Nets 



While this study was vinder way the 

 shrimp fleet, and also the chartered ves- 

 sel, was in the process of changing from 

 the use of a single trawl to a double rig. 

 To determine whether this had an effect 

 on the sampling, the variation of mean 

 lengths and sex ratios of shrimp among 

 samples was analyzed. This also pre- 

 sented an opportunity to check the trawl 

 as a sampling device. 



Repeated samples were taken from a 

 single catch. During cruise 16, the catch 

 from tow 10 was sampled three times 

 and during cruise 17, the catch from tow 8 

 was sampled four times. The average 

 carapace length of males, of females, and 

 of sexes combined are recorded in table 2. 

 The numbers of individuals measured and 

 the sex ratios are shown for each sample. 



