Polovina and Mitchum: Recruitment of Panulirus rmrginstus relative to sea level in NW Hawaiian Is 487 



as variation in tlie SCC, we fo- 

 cused on the analysis of sea level 

 data from the NWHI. Our choice 

 of sea level was primarily a prac- 

 tical one. In comparison to cur- 

 rent or upper-layer temperature 

 records, the sea level records are 

 of long duration, and the data are 

 measured continuously and are 

 available in nearly real-time. An 

 additional advantage is that sea- 

 surface height data from the 

 Geosat satellite altimeter are 

 available to provide a spatial de- 

 scription that complements the 

 temporal description available 

 from the sea level stations. 



Data on the difference in sea 

 level between the gauges at 

 French Frigate Shoals (FFS) and 

 at Midway Is. have beer avail- 

 able since 1976 (Figs. 1,3). This 

 sea level difference (denoted as 

 FFS-Midway sea level) serves as 

 an index of the geostrophic cur- 

 rent anomalies across the NWHI in the region of Maro 

 Reef. For example, an increase in the sea level height 

 at FFS relative to Midway Is., measured from tide 

 gauges, indicates the strengthening of a current that 

 is across the gradient between the two locations and 

 is flowing from the southwest to the northeast. 



To interpret these flow anomalies as a manifestation 

 of the variations in SCC strength, the spatial structure 

 of the sea-surface height variation was examined by 

 mapping the variability observed by the Geosat altim- 

 eter during November 1986-November 1988. These 2 

 years were selected because more accurate orbit 

 estimates were available during this time-period and 

 would result in more accurate sea-surface height 

 fields. The Geosat geophysical data records were ob- 

 tained from NOAA (Cheney et al. 1987) and were pro- 

 cessed with software developed at the University of 

 Hawaii. 



Averages of the Geosat data over November 1986- 

 November 1987 were subtracted from the averages 

 over November 1987-November 1988. Before using 

 the Geosat data, we checked that the resulting sea level 

 differences from the altimeter were consistent with the 

 corresponding sea level differences from tide gauges 

 at FFS and Midway (not shown). Choosing these time- 

 periods also allowed us to contrast conditions during 

 the ENSO period of 1986-87, when the FFS-Midway 

 sea level was low (~520mm), with conditions during 

 the normal period of 1987-88, when the FFS-Midway 

 sea level was higher (~600mm). 



Puerulus settlement 



During the last planktonic stage (i.e., postlarval or 

 puerulus stage), spiny lobster acquire the benthic mor- 

 phological features of adults and become active swim- 

 mers seeking benthic habitat. MacDonald (1986) 

 studied puerulus settlement in the Hawaiian Archi- 

 pelago with traps known as Witham Collectors at Kure 

 Atoll (north of Midway Is.) in 1979-83 and at FFS in 

 1981-85. He computed mean catch per collector over 

 12-month periods (June-May) at Kure Atoll and FFS. 

 These data will be compared with the FFS-Midway sea 

 level data. 



Results 



The fit of the model to the commercial CPUE data and 

 the resulting residuals indicate the model fits the trend 

 in CPUE, but considerable unexplained variation ex- 

 ists in CPUE within and between years (Fig. 4). For 

 example, given the fishing effort, CPUE was greater 

 than expected in 1988 but declined more than expected 

 in 1990. Since the model assumes both constant recruit- 

 ment and constant catchability, the residuals may 

 reflect variation in these factors. From the fit of the 

 model, R = 1.2 xlO^ adult lobsters/yr, M = 0.71/yr, qi = 

 1.2x10-6, and q2 = 0.6xl0-6. Thus 1.2 million lob- 

 sters recruit to the fishery annually; with an M of 

 0.71/yr, only 50% of the 3-year-olds survive 1 year (in 



