tows were shortened. All samples were collected during day- 

 light between 0800 and 1600 h since previous investigations 

 demonstrated that larval abundance is highest at the surface 

 during daylight (Templeman 1937; Templeman and Tibbo 

 1945; Scarratt 1973). The distance covered by the vessel during 

 each tow was determined from loran-C coordinates recorded 

 to the nearest 0.1/js at the start and end of each tow. Surface 

 water temperature was recorded at each station. 



Nets were washed down with seawater after each tow to 

 maintain a standard filtration efficiency. Samples were rinsed 

 in seawater, strained using a 505 \x mesh screen, and placed 

 in buffered 10% Formalin. 8 Samples were sorted for decapod 

 larvae and fish eggs and larvae in the laboratory. All lobster 

 larvae were removed and identified to stage (Herrick 1911) 

 using a dissecting microscope equipped with an ocular micro- 

 meter. Since fourth and fifth stage larvae are best differenti- 

 ated by size (Herrick 1911), larvae identified as fourth stage 

 were measured (total length) to the nearest millimeter and 

 compared with published records of length of stage IV larvae. 

 Larval density estimates were derived using A-distribution 

 theory (Aitchison 1955; Aitchison and Brown 1957), a tech- 

 nique in which survey data are classified into zero and non- 

 zero catch values. The conditional distribution of the non-zero 

 class is assumed to be log-normal. A minimum variance un- 

 biased estimator of the sample mean (Aitchison 1955) is: 



C = m ey Wm (s-/2) 

 n 



where in is the number of non-zero observations, n is the total 

 number of observations, y and s 2 are the mean and variance of 

 the log transformed non-zero observations and ipm (Aitchison 

 1955) is given by: 



Vm = 1 + 



/ + Z 



(«-I) 



j = 2 n J (n+\)(n + ?>). . . (« + 2/-3) j\ 



"Reference lo irade names does noi imply endorsemeni by the National 

 Marine Fisheries Service, NOAA. 



Density estimates for each stage were corrected for stage 

 duration by dividing by development time at prevailing water 

 temperatures (Templeman 1937) for each sample period. Daily 

 production estimates were summed over the entire sampling 

 season to provide an estimate of annual larval production per 

 1,000 m 3 . 



RESULTS AND DISCUSSION 



Seasonal Distribution 



Due to the relatively short duration of the first larval stage, 

 the occurrence of stage I lobster larvae is indicative of recent 

 hatching activity. We initially noted stage I larvae on 1 June 

 and first stage larvae were observed through the beginning of 

 August, indicating that the hatching period in 1978 spanned 

 approximately 2 mo (Fig. 2). Peak abundance of first stage 

 larvae occurred in late June at a mean density of 10.82 larvae/ 

 1,000 m 3 . The mean seasonal density of stage I larvae was 

 1.98/1,000 m 3 . 



Second and third stage larvae were first observed in the 20 

 June collections and the highest densities for both stages were 

 noted in samples collected in late June (Fig. 2). Stage II larvae 

 were collected through the end of July while third stage larvae 

 were observed through mid-August. The mean seasonal den- 

 sity was 3.11/1,000 m 3 and 1.80/1,000 m 3 for second and 

 third stage larvae, respectively. 



Fourth stage larvae were initially collected in the 27 June 

 samples and were dominant for the remainder of the season. 

 The highest mean weekly density of stage IV larvae (26.08/ 

 1,000 m 3 ) was noted in mid-July and fourth stage larvae were 

 collected through the end of August (Fig. 2). Of the 1,030 

 larvae collected during this survey, 645 (62.6°7o) were fourth 

 stage. In contrast, the first through third larval stages com- 

 prised 12.3%, 14.1%, and 10.9°7o of the samples, respectively. 

 The dominance of fourth stage larvae is particularly striking 

 since previous investigators observed relatively low densities 

 for this stage (Templeman 1937; Templeman and Tibbo 1945; 

 Scarratt 1964, 1973; Sherman and Lewis 1967). Bibb et al. 

 (1983) also noted a high proportion (52%) of stage IV larvae in 

 1978 in Block Island Sound, in marked contrast to their 1977 



30 



E 

 o 



o 20 



ac 



Ll) 

 00 



o I 



°n LARVAL STAGE 



Am 



8 19 23 1 6 20 27 5 12 21 28 1 11 15 15 1 



MAY JUNE JULY AUG SEPT 



Figure 2.— Density (no.. 1.000 m') of stage l-IV lobster larvae during sampling period (8 May through 1 September). 



24 



