CHARLES-DOMINIQUE: CATCH EFFICIENCIES OF PURSE AND BEACH SEINES 



marked-2 ones, even though they spent more 

 time inside the beach seine and thus had more 

 opportunities to escape. To explain this result, 

 consider how the two sets of marked fish might 

 have differed: 1) by sampled sizes, 2) by the 

 type of mark, and 3) by the duration of the 

 holding period. Sampling of sizes does not need 

 to be examined because there was no significant 

 correlation between the size and the retention 

 rate. The type of mark itself did not seem likely 

 to influence the fish behavior. However, the dur- 

 ation of the holding period was much longer for 

 marked-1 fish, increasing the opportunities to 

 escape and thus to overestimate Wi. For T. 

 gHineeusis, the two q estimations (Tables 6, 7) 

 differed to a large degi'ee. This was probably due 

 to the stress on marked-1 fish, leading to an 

 overestimation of Wi. and thus to an overestima- 

 tion of the catchability (q,.) calculated by the cap- 

 ture method. The W value that would produce a 

 catchability estimate of q,. = l.'S99( has been cal- 

 culated iteratively and equals 14%. This reten- 

 tion rate is very low but is consistent with the 

 known behavior of this cichlid species, which 

 escapes from beach seines by shpping under the 

 lead line and by jumping over the net (to recover 

 the jumping fish, local fishermen often place 

 small canoes equipped with net curtains along 

 the seine). 



In any case, a retention rate estimated with a 

 marking procedure is greater or equal to the 

 actual efficiency and can be used as an upper 

 estimate of the efficiency. For example, effi- 

 ciency of the purse seine for the crab C. 

 amnicola is smaller than the observed 10% re- 

 tention rate (Table 3). 



The comparison of the retention rates for both 

 seines indicates that the purse seine is more effi- 

 cient in limiting escapement than the beach 

 seine. This can be explained to some extent by 

 the difference in the duration of the sets (22 

 minutes versus 6 hours). The rigging of the 

 gears may also have an influence; the purse seine 

 lead-line hugs the bottom, owing to the drag and 

 weight of the rings, more efficiently than the 

 beach seine; noise and vibrations in the ropes 

 also generally keep the fish away from the net 

 (Hemmings 1967). Therefore, pursing is more 

 efficient than the manual closing of the beach 

 seine. The better efficiency of the purse seine 

 should, however, not be generalized because the 

 efficiency of a particular gear may be influenced 

 by subtle differences of rigging (MacMullen 

 1981). We did observe during another experi- 

 ment'' some important differences in the efficien- 



cies of two apparently similar purse seines, prob- 

 ably resulting from a slight difference in lead- 

 line weights. 



Species and Size Selectivity 



Other robust results came from a comparison 

 of the catch rates of the two gears in terms of 

 species relative abundance (species selectivity) 

 and size distribution (size selectivity). 



Species selectivity is due to differences be- 

 tween the efficiencies, which depend upon com- 

 plex interactions between the gears and the be- 

 havior of the species. In this experiment, an 

 additional "enclosure effect" can happen if all fish 

 are not equally available. For instance, by 

 crowding along the net, the fish become inacces- 

 sible to the purse seine {p_.x, the overall accessi- 

 bihty is then less than 1). 



The ratio of catch rates by the two gears, 

 calculated in Table 8, can be compared to the 

 theoretical value that would be obtained if both 

 gears were equally efficient, and if the avoidance 

 rate for beach seine was negligible. This can be 

 assumed as a first approximation since the net is 

 very large (1,100 m) and fairly silent (no engine 

 was used in the boat). 



From the formulas given above (see methods: 

 estimation from captures): 



q-C 



1 



and *■ = — -a, 



W (1 = qf-' W 



with a = 1 ; -H . . . -I- 



(1 - q) (1 - q)- 



(1 + qf 



since q = p„  e, then r = Pa  (e/W)  a. 



The parameter )• appears to be roughly propor- 

 tional to e/W, a being a correction factor account- 

 ing for the successive catches in the enclosure. 

 The parameter a, depending on the value of e, 

 which varies between and 1, is in the interval 

 (1-1.276). If the efficiencies of the two gears are 

 equal {e = E or e = W, since U is assumed to be 

 equal to 1), r is in the interval [p„-1.276 • p„], 

 that is [0.076-0.097]). In Table 8, r is smaller 

 than 0.076 for most fish species, indicating lower 

 efficiency of the purse seine. On the contrary, 



■'Cantrelle, I., E. Charles-Dominique, Y. N. N'Goran, and 

 J. Quensiere. 1983. Etude experimentale de la selec- 

 tivite de deux sennes tournantes et coulissantes (maillage 

 25 mm et maillage mixta 14-25 mm) en lagune Aby (Cote 

 d'l voire). Unpubl. rep., 36 p. Cent. Rech. Oceanogr, 

 Abidjan. 



919 



