PARRACK: FISHING EFFORT FROM AERIAL SEARCH DATA 



which any fishing occurred, an observation of a 

 vessel fishing was logically defined as an observed 

 day fished. Although the majority of vessels 

 sighted were in some phase of fishing operations, 

 some vessels were observed to be in other opera- 

 tional modes such as drifting, steaming, anchored, 

 loading, unloading, or jogging in heavy seas. 



These observations therefore allowed certain 

 fishing effort variables to be derived by nationality 

 and gear type. They include: 1) the number of 

 times any vessel was observed on the grounds (i.e., 

 observed days on grounds), 2) the number of times 

 those same vessels were observed fishing (i.e., 

 observed days fished), and 3) the number of 

 vessels. 



These overflight data were subject to limita- 

 tions which were accounted for in the analysis. 

 First, some observed vessel days have probably 

 been incorrectly categorized. Surveillance flights 

 usually occurred before midday; consequently a 

 sighted vessel that did not fish until late in the day 

 was recorded as not fishing. Such an event was 

 therefore interpreted in the analysis as an ob- 

 served day on grounds but not as an observed day 

 fished as would have actually been the case. This 

 limitation, as will be explained later, has little 

 effect on the estimation of days fished if such 

 inaccuracies are constant in magnitude through 

 time. These data were further limited in that 

 incorrect vessel identifications sometimes oc- 

 curred. Adverse weather conditions, dense fleet 

 concentrations, hull scripts of poor visibility, and 

 inaccurate interpretations of non-Roman script 

 resulted in the recording of incorrect individual 

 vessel identifications. Lists received from certain 

 countries (Japan, Romania, Spain) made possible 

 the verification of hull scripts observed during 

 1974. The comparison of these reported hull iden- 

 tifications with those recorded on overflights 

 during 1974 determined that individual vessel 

 identifications recorded two or more times on 

 fisheries surveillance operations were almost al- 

 ways correct; those recorded only once were almost 

 always incorrect. For example, 40 Spanish stern 

 trawlers were in ICNAF Subarea 5 and Subarea 6 

 (Figure 1) during the first 10 mo of 1974^. During 

 that period, 39 separate Spanish stern trawlers 

 were recorded on fisheries surveillance activities 

 more than once. The number of vessels in the area 



during a time period of interest was therefore 

 established by considering only vessel identifica- 

 tions observed by fisheries surveillance personnel 

 more than once over the period 1965-74. 



Knowledge of each country's fishing effort, both 

 total and by types of gear used, and of the total 

 effort expended by all countries, is of prime con- 

 cern in existing fisheries management regimes. 

 Separate estimates were therefore made for each 

 country and vessel-type, as were estimates of each 

 country's total effort and estimates for each of the 

 total stern trawl and total side trawl fleets. 



In addition it was hypothesized that the relation 

 between reported and observed effort for the 

 various gear and nationality components could be 

 different. Stern trawlers are larger and were 

 expected to be of greater visibility than smaller 

 vessels. Also, surveillance searching operations 

 were likely directed towards certain national 

 fleets as a result of their greater size, their pres- 

 ence in an area closed to fishing, or because their 

 catches were of particular immediate concern. If 

 such relationships are different, separate es- 

 timates of functions of sighted and reported effort 

 for each fleet component would logically increase 

 estimation accuracy. 



Estimator I 



The ratio of reported days fished to the average 

 number of sighted days on grounds per flight were 

 easily computed for time periods when reported 

 fishing effort was available: 



R=f/g' (1.1) 



where 



r = (t ^.) -. 



^ICNAF. 1974. Comments of the Spanish delegation on the 

 U.S. memorandum annexed to ICNAF Comm. Doc. 74/41. Special 

 Meeting ICNAF Comm. Doc. 74/44, Ser. No. 3422. 



A = the number of flights made during the time 



period, 

 g ' = the number of sighted days on grounds 

 during the ? th flight, and 

 / = the number of days fished during the time 

 period as reported to ICNAF. 



This ratio may then be applied to aerial observa- 

 tions in some future time period to estimate days 

 fished before the value is reported: 



f=Rg'. (1.2) 



R is computed from previous data and g' is cal- 



505 



