492 



Fishery Bulletin 90(3). 1992 



The underlying mechanism Unking the correlation 

 between the SCC and subsequent recruitment at Maro 

 Reef is not known. It is possible that the SCC returns 

 larvae, which have been advected west of the archi- 

 pelago, back to Maro Reef. The SCC has been hypoth- 

 esized to transport Acropora coral from Johnston Atoll 

 (lat. 16°45'N., long. 169°31'W) to FFS (Grigg 1982). 

 In addition, larvae of a spiny lobster species not re- 

 corded as an adult in Hawaii have been transported 

 from the Marshall Is. to the Hawaiian Archipelago 

 (Phillips and McWilliam 1989). 



However, it may be that the SCC impacts not advec- 

 tion but larval survival. Laboratory studies have shown 

 that spiny lobster larvae suffer a high level of mortal- 

 ity when water temperatures drop below 20°C (T. 

 Kazama, NMFS Honolulu Lab., pers. commun., Sept. 

 1991). In the years we estimated that the SCC was 

 weak, water temperatures <20°C in the winter have 

 been observed at Maro Reef but not Necker I. If little 

 larval mixing occurs between Maro Reef and Necker 

 I., larval mortality at Maro Reef resulting from low 

 winter temperatures could account for the observed 

 recruitment variation. 



A third hypothesis is that when the SCC has a par- 

 ticular speed and location, it produces fronts which re- 

 tain larvae near Maro Reef. When the SCC is weak or 

 shifts, these fronts are not formed near Maro Reef. 

 Preliminary evidence from the drifter buoys and lar- 

 val sampling in our study suggests fronts north of Maro 

 Reef and south of Necker I. may be important for 

 lobster larvae (Polovina, pers. observ.) 



o 

 "5 



"5 



x> 

 _o 



c 

 a. 



1234123412341234123412341234123412341234 



Quarter of year 



Figure 12 



Three-quarter moving average of the ratio of quarterly landings of spiny lobster PanuLirus 

 marginatus at Maro Reef to quarterly landings at Maro Reef and Necker I. (bold line) 

 overlayed with the 3-quarter moving average of French Frigate Shoals (FFS)-Midway 

 sea level deviation advanced by 4 years (thin line). 



One potential management application of the lagged 

 relationship between FFS-Midway sea level and recruit- 

 ment is that it provides up to a 4-year forecast of re- 

 cruitment to the fishery at Maro Reef. A 3-quarter 

 moving average of the FFS-Midway sea level shifted 

 forward by 4 years forecasts poor recruitment in 1991, 

 followed by an improvement beginning in late 1992 (Fig. 

 12). During January-May 1991 before the fishery was 

 closed for 6 months, recruitment at Maro Reef clearly 

 had not recovered, as only 1052 spiny lobster were 

 harvested from Maro Reef while 34,746 spiny lobster 

 were harvested from Necker I. Recall that the relative 

 catches between banks provide an index of relative abun- 

 dance, since the fleet moves to maximize the CPUE. 

 The FFS-Midway sea level data forecast that catches 

 at Maro Reef will improve beginning in late 1992 (Fig. 

 12). Data from larval tows are consistent with this 

 forecast. Standardized larval tows, taken in June and 

 November 1989 over a grid of stations from the 200 m 

 isobath out to 56km around both Necker I. and Maro 

 Reef, caught 3802 and 3342 late-stage phyllosomes, 

 respectively (J. Polovina, unpubl. data). A f-test, based 

 on a lognormal distribution, finds no significant differ- 

 ence in the mean abundance of larvae between Maro 

 Reef and Necker I. If we assume that larval abundance 

 was high around Necker I. in 1989, then good larval 

 recruitment apparently has returned to Maro Reef. 

 This is consistent with the observed higher sea-level 

 values in 1989 (shown as 1993 values in Fig. 12, since 

 the sea level has been advanced by 4 years) and sug- 

 gests that catches will be high at Maro Reef in 1993. 

 The FFS-Midway sea level time 

 series from 1976 to 1990 (Fig. 3) 

 shows that ENSO events may re- 

 sult in poor recruitment to the 

 fishery 4 years later, but the series 

 also shows a long-term decline. 

 Reasons for the low FFS-Midway 

 sea level during ENSO events are 

 not known, but may be related to 

 a decrease in surface water sup- 

 plied to the SCC in the western 

 Pacific. Such a change could be 

 associated with the circulation 

 disruptions observed in the trop- 

 ical Pacific during ENSO events 

 (Meyers and Donguy 1984). The 

 long-term decline in sea level 

 from 1976 to 1990 suggests there 

 is a low-frequency component in 

 the variation in SCC strength 

 and, hence, lobster recruitment. 

 Thus, it may be some time before 

 recruitment to the fishery is at 

 the early 1980s' level. 



800 



750 



700 



650 



E 

 E 



01 



> 



0) 



a 



0) 



o 



T3 



I 



in 



