Dentzau and Chittenden: Population dynamics of Polydactylus octonemus in the Gulf of Mexico 



455 



m deep in March over the West Flower Garden Bank 

 off Texas (27°52'N, 93°48'W; Bright and Cashman 

 1974), (c) 50 mm in water < 22 m in April off Barataria 

 Bay, LA (Gunter 1938b), and (d) 56-66 mm in water 

 54 m deep in early May off Freeport (M. Rockett, Ap- 

 palachian State Univ., Boone, NC 28608, pers. com- 

 mun., March 1981); (3) collection at the surface in early 

 May of "young"^ off south Texas (26°05'N, 95°25'W) 

 and off the Louisiana continental shelf (26°40'N, 

 92°00'W) in water 1890 and 1602 m deep, respective- 

 ly (Table 6; Springer and Bullis 1956); and (4) collec- 

 tion of neustonic "larvae" (footnote 1) 65-72 mm in 

 water 91-134 m deep in mid-April-mid-May off south 

 Texas (Wormuth et al. 1979). 



Our calculated spawning-period duration of 45-120 

 days assumes large fish hatch before small ones and 

 all grow at the same rate (Geoghegan and Chittenden 

 1982). The latter assumption appears reasonable, be- 

 cause 99% confidence intervals for observations (in 

 Dentzau 1985, table 3) were fairly constant between 

 cruises within each May-June period. Our calculated 

 duration may be an overestimate because continued 

 recruitment of small fish to the Gulf would tend to 

 depress estimates of mean growth/day, the denomin- 

 ator in our spawning-duration equation. The numer- 

 ator, mean size range, is less affected; larger recruits 

 gradually disperse offshore to deeper waters, but 

 we collected throughout and beyond the bathymetric 

 range of the species. Comparisons of spawning-period 

 durations from 90 and 99% confidence intervals are 

 valid, however; such calculations use the same values 

 except t. 



Our calculated hatching dates indicate mid-January- 

 mid-February spawning, these being mean spawning 

 dates because regression predicts averages. They are 

 in error to the extent that regression curves do not pass 

 through the origin and that they are asymptotic to the 

 X- or ^/-axis in very early life. The latter is an unknown; 

 the former problem does not seem serious, because 

 Shirota (1970, in Hunter 1981) found length at first 

 feeding in many species is four times the egg diameter. 

 Egg diameter is not known in P. octonemus (Martin and 

 Drewry 1978, deSylva 1984), but it averages 0.76 mm 

 in the closely related Eleutheronema tetradadylum 

 (deSylva 1984). This suggests P. octonemus is 3-3.5 mm 

 at hatching, so only a few days error occurs in x, at 

 most. 



Spawning areas 



Any spawTiing of P. octonemus in the northern or north- 

 western Gulf presumably occurs in the water column. 



■•Springer and Bullis (1956) defined young as postlarval. juvenile, and 

 young specimens. 



Our data and the literature (Chittenden and McEach- 

 ran 1976; Wohlschlag et al. 1977, 1979) indicate they 

 basically disappear as demersal fish from the white and 

 brown shrimp communities of the northern Gulf in 

 November-December. They may assume a pelagic 

 behavior then, which would explain their apparent 

 absence in January-February, because Hastings et al. 

 (1976) observed this species in December in habitat 

 typical of pelagic fish— open water areas around an off- 

 shore platform in water 18 m deep in the northeastern 

 Gulf off Panama City, Florida. 



Catch records for young in the literature and the 

 disappearance of P. octonemus from the white and 

 brown shrimp communities in their primary winter 

 spawning period, if the latter does not simply reflect 

 gear selectivity following assumption of a pelagic habit, 

 suggest this species spawns along the outer continen- 

 tal shelf, the continental slope, or further offshore. 

 Although pelagic "larvae" (footnote 1) and juveniles 

 have been collected at the surface in or near waters 

 of the white shrimp community off Texas and Louisiana 

 in April-early June (collections 1 and 12, Fig. 13) or 

 the brown shrimp community in late February-May 

 (collections 8, 10, 15, 16, and 23, Fig. 13), more fre- 

 quent collections in U.S. waters have been made along 

 the outer continental shelf at 106-137 m in January- 

 May (collections 5, 6, 7, 9, 11, 13, and 14, Fig. 13), and 

 even further offshore at > 200-2300 m in March-May 

 (collections 2, 3, 4, 18, 19, 20, 21, and 22, Fig. 13). 

 Several collections of "young" (footnote 4) and juve- 

 niles have been made in Mexican waters 1035-2736 m 

 deep between Laguna Madre Tamaulipas and Cam- 

 peche Bank in March-May (collections 24-32, Fig. 13). 

 Similarly, young of other Polydactylus spp. also occur 

 off continental shelves although adults are common in 

 inshore waters; e.g., Klawe and Alverson (1964) found 

 young P. opercularis and P. approximans <48 mm 250 

 nm offshore in the eastern tropical Pacific Ocean. 



In contrast to the abundance of P. octonemus larvae 

 and juveniles in the western and northwestern Gulf, 

 they have not been identified in extensive collections 

 off the west coast of Florida from Pensacola on the 

 north, south to the Florida Keys (Vick 1964; Houde 

 et al. 1979; M. Lieby, Bur. Mar. Res., St. Petersburg, 

 FL 33701, pers. commun., Feb. 1986). This suggests 

 little or no spawning in that area and in the eastern 

 side of the Loop Current, which contributes water to 

 the western Florida shelf (Austin and Jones 1974). 



Polydactylus octonemus eggs, larvae, and juveniles, 

 pelagic like other polynemids (Breder and Rosen 1966, 

 Kagawade 1970), presumably use current transport to 

 reach their estuarine and shallow Gulf nurseries. Stan- 

 dard and Chittenden (1984), based on Murphy and Chit- 

 tenden (unpubl.), suggested spawning in Larimusfas- 

 ciatus and other Gulf fishes is timed to coincide with 



