FISHERY BULLETIN: VOL 77, NO 1 



explain the observed distribution of flotsam- 

 associated sets. 



The average currents in the eastern tropical 

 Pacific, as derived from ship's drift data, were de- 

 termined by Wyrtki (1965). From January until 

 May, the California Current is strong. Circulation 

 near Area 3 is to the south. Circulation near Areas 

 1 and 2 is gyral. From May to July, both the 

 Equatorial Countercurrent and the California 

 Current are relatively stong. During this period, 

 most countercurrent water turns north and flows 

 along the coast of Central America. Area 3 has a 

 northern and southern flow, the northern flow 

 along the coast. Area 1 maintains its gyral flow. 

 From August through December, the Equatorial 

 Countercurrent is well developed. Circulation in 

 Area 3 is to the south. Area 2 maintains its north- 

 western flow along the coast and Area 1 flow main- 

 tains a gyral pattern. If logs disperse mainly from 

 the Gulf of Nicoya, the Gulf of Tehuantepec. and 

 the Gulf of Fonseca, then the gyral circulation in 

 Area 1 would tend to maintain logs and other 

 flotsam in the area for a considerable time. The 

 northwest coastal current in Area 2 could trans- 

 port flotsam through Area 2 and during part of the 

 year into Area 3. Since the North Equatorial and 

 South Equatorial Currents are rather strong, one 

 would not expect floating objects to persist in Area 

 4 except near the boundaries with Areas 1 and 2. 

 Hence the location of large rivers and the system 

 of currents is reasonably consistent with the geo- 

 graphical distribution of flotsam-associated sets. 



In order to compare different set types, Spear- 

 man's rank correlation coefficient was calculated. 

 For each 5° square in the C YRA, the total numbers 

 of flotsam-associated sets, porpoise-associated 

 sets, and unassociated schoolfish sets were tabu- 

 lated for each year. These totals were ranked and 

 the ranks were correlated. Only 5° squares where 

 at least 10 sets occurred were used in calculating 

 correlations. When a minimum of 40 sets was used 

 as the criterion for including a 5° square, the corre- 

 lations were qualitatively the same as with the 10 

 sets criterion. The results (Table 3) show that a 

 significant positive correlation exists between 

 number of sets on unassociated schoolfish and 

 flotsam-associated tuna. Porpoise sets were uncor- 

 rected with other set types. 



The above results indicate that fish caught as- 

 sociated with flotsam tended to be caught in the 

 same area at the same time as unassociated school 

 fish. Examination of available length-frequency 

 data on a species basis, weighted by the catch in 



T.^BLE 3. — Spearmans rank correlation between three types of 

 sets by year. Number of sets/5° square. (Source of data: I ATTC . ) 



each set (Figure 3), indicated that unassociated 

 schoolfish and flotsam-associated yellowfin and 

 skipjack tuna had very similar length-frequency 

 distributions. The length-frequency information 

 and the correlation analysis support the idea that 

 unassociated tuna and flotsam-associated tuna 

 are related. Flotsam, acting as an attractant, may 

 aggi-egate tuna that would otherwise be caught in 

 unassociated sets. 



The number of flotsam-associated sets has in- 

 creased dramatically since 1971 (Figure 2). The 

 trend in percentage of flotsam-associated sets (Fig- 

 ure 4) indicates that flotsam-associated sets have 

 increased in importance to the fishery. Stratifying 

 the number of unregulated flotsam-associated sets 

 by area (Figure 5i shows that the trend of more 

 flotsam-associated sets is not an area effect. All 

 areas, except Area 3, have shown a marked in- 

 crease in number of flotsam-associated sets. Area 

 3 does not show an increase because logs are only 

 deposited in this region during a limited portion of 

 the year. In January-May, the near surface cur- 

 rent in Area 3 is to the south (Wyrtki 1965), cut- 

 ting off the source of logs that wash down the 

 rivers of Central America. Also, good fishing often 

 occurs in Area 3 during the later months of the 

 year, a period not included in my unregulated 

 data. It appears that the increase in flotsam- 

 associated sets in recent years was not caused by 

 discovery of new areas with abundant flotsam but 

 rather by an increase in fishing effort on flotsam in 

 all areas but Area 3. 



Average rainfall in Central America was tabu- 

 lated (Table 4) to see if there was a correlation 

 between river runoff and the number of flotsam- 

 associated sets. Comparison of number of 

 flotsam-associated sets and rainfall revealed only 



150 



