528 



Fishery Bulletin 97(3), 1999 



distribution of size cohorts and the occurrence of 

 ovigerous females in the population. Temporal trends 

 in both carapace lengths and sex ratios for the over- 

 all four-year span of the study were analyzed by lin- 

 ear-linear piecewise polynomial regression to define 

 slope transition points. Percentile data for the sex 

 ratios were arcsine-transformed prior to these analy- 

 ses. In order to remove effects of morphologically 

 anomalous postreproductive and senescent individu- 

 als, animals >18.0mm CL (two females and three 

 males) were excluded from analyses of relative 

 growth. Linear-linear piecewise polynomial analy- 

 sis of untransformed data was used to optimize posi- 

 tioning of ontogenetic transition points (±0.05 mm) 

 in all relative growth comparisons. Except for this 

 alternative method of iteratively locating slope tran- 

 sitions, methods complied with the recommended 

 practice of fitting regi'essions to untransformed data 

 above and below transition points by reduced major 

 axis (Lovett and Felder, 1989). Standard allometric 

 coefficients (Huxley, 1932 ) were also determined from 

 regressions of log- transformed data by least squares 

 estimate, with data subdivided at the previously es- 

 timated transition points. All statistical analyses 

 were performed with NCSS® (Number Cruncher 

 Statistical System) 6.0 software (Hintz, 1995). 



Because generic revisions have yet to address many 

 of the confamilial taxa that are treated in our com- 

 parative discussions, it was necessary in compara- 

 tive discussions to refer to some species under 

 Callianassa sensu lato (s.l.), while acknowledging 

 that Callianassa sensu stricto has been restricted to 

 a few eastern Atlantic populations (see Manning and 

 Felder, 1991). 



Results 



Sex ratio 



Except for January to June 1992, March 1993, and 

 February 1995, sex ratios were significantly female- 

 biased over the four years of sampling (Fig. lA). Over 

 the entire study period, the mean sex ratio was 2.4 

 +0.3 females per male, with the lowest mean ratio 

 (1.1) occurring in February 1992, and the highest 

 ratio (5.1) in October 1993. Trends in mean CL of 

 the population (Fig. IB) generally tracked those in 

 sex ratios, as defined by piecewise regression analy- 

 sis. Slopes over the first two years revealed declin- 

 ing relative abundance and increasing mean CL of 

 males, patterns that differed significantly from those 

 of the last two years of monitoring. An iteratively 

 optimized transition point for sex ratios occurred at 

 23 months, whereafter mean sex ratios remained 



near asymptotic. Analysis over the full study period 

 also defined an optimized transition point or peak in 

 CL of males at 23 months, beyond which male CL 

 was asymptotic. The CL of females, however, peaked 

 late in the first year of monitoring. 



Mean size of both males and females increased 

 slightly but significantly from the first through fourth 

 quarter of each year except for 1994, a year in which 

 the presence of first quarter recruits was offset by a 

 cohort of large individuals surviving from the previ- 

 ous year (Fig. 2). When treated separately, mature- 

 sized individuals (>11 mm CL) of both male and fe- 

 male populations increased in mean CL over the year, 

 usually peaking in the fourth quarter prior to maxi- 

 mum abundance of ovigerous females in the follow- 

 ing year (Fig. 3, B and D). Mean CL of males ap- 

 peared to exceed slightly that of females over the full 

 study period, and males dominated large size classes 

 (Figs. IB, 2, 3B). Only 4.37f of the collected females 

 were >15.0 mm CL, whereas 12.4% of males attained 

 this size. The largest animal collected was a 20.7- 

 mm-CL male. Detection of recruits <8-10 mm CL 

 was sporadic throughout the study (Fig. 3E), and 

 tracking of growth in recruitment cohorts was not 

 possible. 



Ovigerous females 



Among all ovigerous females collected during the 

 study (/!=444), size ranged from 7.04 mm CL to 16.8 

 mm CL. However, 38.3'7( of all these were in the 13.0 

 to 13.9 mm CL size class and <1.59^ of ovigerous fe- 

 males were <11.0 mm CL (Fig. 4). The single oviger- 

 ous specimen <9.8 mm CL was an apparently preco- 

 cious individual taken in August 1994, while the 

 population of ovigerous females was low (Fig. 3D). 

 During peak abundances of ovigerous females (Feb- 

 ruary to June, Fig. 5B), 64.3'/f of females in the 13.0 

 to 13.9 mm CL size class were ovigerous (Fig. 2). 



Ovigerous females were found in all months of the 

 four-year study (Figs. 2 and 3D). However, percent- 

 ages were lowest in the third quarters of 1992, 1993, 

 and 1995, and the fourth quarter of 1994, whereas 

 the highest percentages occurred in the first or sec- 

 ond quarter of each year. Peak abundance of oviger- 

 ous females coincided with the highest quarterly sa- 

 hnity in 1992, 1994, and 1995 but was less defined 

 in 1993 when the fourth quarter salinity did not fall 

 toa typical low value (Fig. 3, A and D).When monthly 

 samples were pooled over the entire study, highest 

 percentages of ovigerous individuals among all fe- 

 males were evident during the high salinity period 

 from February to June; a maximum was found in 

 May (35.7%), three months after peak development 

 of the ovaries (Fig. 5, A-C). 



