establish an experimental population from which 

 could be obtained measures of growth and mor- 

 tality in the parent population before and during 

 as well as after its transition from prerecruit to 

 postrecruit status. 



Accordingly, operations using facilities provided 

 by the Bureau's M/V George M. Bowers got under- 

 way on September 18, 1961, with the first releases 

 of marked shrimp being made 2 days later. 

 Shrimp were captured with small-mesh trawls 

 hauled for very short intervals of time (frequently 

 not more than 2 minutes) to minimize injury 

 caused by compression in the net. To insure that 

 only shrimp in prime condition entered the experi- 

 ment, catches were held in tanks of circulating sea 

 water for periods of at least 12 hours. After all 

 weak and dying individuals had been removed, the 

 remainder was sorted so that only those falling 

 into a specified size range (total length) were 

 retained for marking. This range was arbitrarily 

 set using as a point of reference the modal length 

 of available shrimp, such length having been 

 determined by sampling catches made at the start 

 of operations. Graded individuals were then 

 marked by carefully injecting small quantities of a 

 0.25-percent aqueous solution of Trypan blue stain. 

 Groups of marked shrimp were observed for at 

 least 4 hours before being released, only the most 

 vigorous individuals at the end of that time being 

 selected for the experiment. These were released 

 in lots of 100 at scattered sites and at irregularly 

 spaced intervals during the marking period by 

 means of a special release box (Costello, 1964). 

 This device permitted the experimental shrimp 

 to be returned directly to the bottom from whence 

 they came (the depth being about 12 fathoms), 

 and precluded large-scale losses due to predation 

 by fishes and birds that would have occurred had 

 the shrimp been released at the surface. Under- 

 water movies taken during this operation reveal 

 that marked shrimp released from the box immedi- 

 ately burrowed into the bottom, thereby satisfying 

 the assumption that predation losses were, for all 

 practical purposes, nonexistent. Because of judi- 

 cious handling, selection, and release of the experi- 

 mental material, it is hereinafter assumed that the 

 total number of marked shrimp reintroduced into 

 the parent population represented the actual or 

 "effective" size of the marked population estab- 

 lished during the marking period. 



In the manner just described, an experimental 



population of 2,090 individuals was created within 

 a 4-day period. Figure 1 shows the general loca- 

 tion of capture and release sites. Measurements 

 from an accumulation of small samples drawn 

 each time a group of shrimp was released provided 

 essential information concerning the population's 

 length and weight attributes at the start of the 

 experiment (table 1). 



COMMERCIAL FISHING OPERATIONS DURING 

 EXPERIMENT 



Greatly influencing the outcome as well as the 

 utility of results of experiments like the present 

 one are the amount and distribution of recapture 

 effort relative to the distribution of the experi- 

 mental population. With all other potential 

 sources of bias inoperative or satisfactorily ac- 

 counted for, the rate at which marked individuals 

 are recaptured becomes simply a function of the 

 probability of recapture. 3 An index of this 

 probability is, logically, the intensity of fishing 

 or, in a rather restricted area such as is involved 

 here, the fishing effort. A major problem is that 

 the fishing effort often does not remain constant 

 and thereby confounds the probability of re- 

 capture. It must be assumed at the outset, of 

 course, that the experimental population soon 

 reintermingles with and behaves in the same 

 manner as the parent population (age group) from 

 which it was temporarily removed. 



Detailed information of fishing operations during 

 the Tortugas experiment was provided by the 

 Bureau of Commercial Fisheries Branch of 

 Fishery Statistics which has agents stationed at 

 major Gulf ports, including Key West, Fort 



Table 1. — Sex, length, and weight data from samples of the 

 marked population at the beginning of the Tortugas experi- 

 ment, middle of week ending Sept. S3, 1962 

 [Number in marked population: 2,090. Estimated sex ratio: 60^:409] 



' Carapace length is denned as the straight-line distance from the postero- 

 most edge of the orbit to the posterior edge of the carapace where it intersects 

 the iniddorsal line. 



3 Throughout this report, "recapture" refers to the marked shrimps' capture 

 by the fishing gear; in contrast, "recovery" refers to their subsequent detec- 

 tion in commercial catches (or landings) . 



DYNAMICS OF A PENAEID SHRIMP POPULATION 



317 



