collections in which stage IV larvae composed 4% of the total 

 larvae obtained. 



High stage IV densities may be due to transport into the 

 study area or differential catchability. Phototropic responses 

 differ among larval stages (Hadley 1908; Templeman and 

 Tibbo 1945; Ennis 1975) with first and fourth stage larvae 

 being most strongly phototactic, indicating that these stages 

 may be more vulnerable to surface gear. Since previous re- 

 searchers, using similar gear, obtained relatively low stage IV 

 densities, it is unlikely that differences in catchability alone 

 are sufficient to explain the high abundance of fourth stage 

 larvae in our samples. 



100 -u 



50- 



LARVAL STAGE 



I 



12 3 4 5 



buk 



L r° — r — i — ? — 7 



100 -i 



— 50- 



>- 



o 



z 0- 



LU 



o too- 



L±J 



u. 



10 



^ 1 

 12 3 4 5 



i r 



C3 l~l 



I I 



50- 







10 



^ — i 



12 3 4 5 



JK 



^t " r r 



100n 



50- 







20 

 10 



1 2 3 4 5 



IV 



■ i m r — f^^ 



10 20 



DENSITY 



30 



Figure 3. — Frequency distribution of lobster larval catch densities (no. /1, 000 m 2 ) 

 and distribution of log transformed catch densities (zero catches excluded). 



Conditional distribution of the non-zero catches was 

 approximately log-normal for each larval stage (Fig. 3). Con- 

 tagious distribution patterns have been consistly noted in 

 larval lobster sampling programs (see reviews by Fair 1980 and 

 Stasko 1980). Patch size dimensions are not generally known, 

 however no significant differences in catch between paired 

 neuston nets separated by =15 m were observed (Wilcoxon 

 paired rank sum test; P>0.05). 



Comparisons between larval catches at each of the eight 

 stations (Table 1) indicated no significant differences among 

 stations for stages I-III (Kruskal-Wallis test; P>0.05). The 

 lack of significant differences in stage I densities between 

 stations does not permit inference regarding possible spawning 

 locations; however, the area surveyed was relatively small and 

 transport of larvae with prevailing currents may have ob- 

 scurred source areas. Significant differences (P< 0.005) were 

 noted, however, between stations for stage IV larvae (Kruskal- 

 Wallis test; x 2 = 20.91; df = 7). Stage IV densities tended to 

 be highest in the western segment of the study area (Table 1), 

 however, relatively high stage IV density was noted at station 1 

 in the eastern section of the survey area. 



Table 1. — Mean larval densities (no. /1, 000 m 3 ) for 

 stages I-IV over the entire sampling season. 







Larval 



stage 





Station 



I 



II 



III 



IV 



1 



2.208 



4.197 



2.560 



9.852 



2 



1.071 



2.456 



1.995 



2.242 



3 



0.559 



0.938 



0.331 



0.891 



4 



3.585 



9.339 



2.099 



0.699 



5 



3.383 



5.857 



1.035 



12.990 



6 



1.231 



1.633 



3.957 



12.166 



7 



0.207 



0.560 



1.533 



10.816 



8 



0.310 



1.540 



0.790 



1.969 



Seasonal Production 



Seasonal production curves were constructed based on stan- 

 dardized daily production estimates (Fig. 4); annual larval 

 production was then determined by integration. Estimated 

 annual production for stages I-IV was 22.72, 19.80, 12.20, and 

 32.33 larvae/ 1,000 m\ respectively. The high production esti- 

 mate for fourth stage larvae, despite correction for stage dura- 

 tion, is indicative of the unusually high abundance of stage 

 IV larvae in 1978. The adjusted estimates of stage IV produc- 

 tion are conservative since settlement occurs approximately 

 midway through the fourth stage and the larvae are no longer 

 vulnerable to the gear (Scarratt 1973). Scarratt (1964) provided 

 stage 1 seasonal production values for Northumberland Strait 

 which considerably exceeded our estimates. 



Stage I density was expanded to provide an estimate of 

 2.514 x 10 6 stage I larvae produced in Rhode Island Statistical 

 Area 4 (Fig. 1). The statistical area encompasses 165.01 km : 

 and it was assumed that larvae were confined to the upper 0.5 

 m of the water column. The stage I total production estimate 

 was corrected for an instantaneous daily mortality rate of 

 Z = 0.050 derived by regressing log e transformed production 

 of stages I-III on the weighted mean duration (days) of each 

 stage. Due to the many variables influencing the catchability 



25 



