McFARLANP ET AI..: RKCKl'ITMENT PATTERNS IN FRENCH GRUNTS 



specific attempts to validate spawning in French 

 grunts in the sea at St. Croix, two of us (McFarland 

 and Shulman) failed to observe reproductive activity 

 during dusk, but we did observe what appeared to be 

 sporadic spawning by small groups of tomtate 

 grunts, Haemulon aurolineatum, within large 

 schools of these fish. These limited data suggest that 

 grunts, like many reef fishes, cast pelagic eggs into 

 the water column at dusk. We emphasize, however, 

 that there are no data about their daily spawning 

 habits. 



Two hypotheses offer explanations for the domi- 

 nant semilunar periodicity of fertilization and settle- 

 ment. 



Hypothesis 1. Assumption: Spawning follows a 

 semilunar rhythm with breeding peaks closely cou- 

 pled to the quarter moons (and/or intermediate 

 monthly tides). If reproductive activities in French 

 grunts follow a semilunar cycle, and pelagic life is 

 programmed for 15 d, then settlement should occur 

 most often during the quarter moons and interme- 

 diate monthly tides, which it does (see Table 3, Fig. 

 6). This hypothesis, however, does not account for 

 the weekly peaks in settlement and fertilization 

 (Table 3; Figs. 2, 3, 5), which contradict the assump- 

 tion of the hypothesis. 



Hypothesis 2. Assumption: Spawning is relatively 

 constant from day to day, and larval existence 

 restricted to about 15 d. Consistent daily reproduc- 

 tive effort could produce a continuous pool of grunts 

 in the plankton. As a result, only those larvae that 

 are favored by "correct" currents that disperse them 

 inshore around 15 d will settle, and/or successful 

 recruitment may also depend on favorable currents 

 dispersing the eggs at the time of reproduction. At 

 St. Croix we suspect that current conditions are 

 most favorable to settlement during the quarter 

 moons and intermediate tides. The weekly peaks 

 that are associated with full and new moons (Table 3) 

 could represent recruitment in less favorable cur- 

 rents than occur around the quarter moons. These 

 currents, nevertheless, must allow some larv^ae 

 through the "filter screens" that all recruits must 

 pass through to join a reef community (Smith 1978). 

 The semilunar rhythm of fertilization and settlement 

 observed in settling grunts would, under this hypo- 

 thesis, be explained by semilunar rhythms in cur- 

 rents favorable to settlement. 



The time of day when settlement occurs and, espe- 

 cially, the amount of darkness at night may also 

 relate to successful settlement. Although data are 

 scarce, some reef fishes settle from the plankton at 

 night (e.g., acanthurids- Randall 1961; McFarland, 

 unpubl. data). We do not have similar direct observa- 



tions for French grunts, but we suspect settlement is 

 nocturnal because the number of PL-l's remained 

 fairly constant on most census sites throughout each 

 day. Nocturnal settlement behaviors would tempora- 

 rily remove recruits from the attack of reef pisci- 

 vores, especially if accomplished during the dark of 

 the moon. During full moons, recruits presumably 

 would be subject to higher rates of predation by 

 planktivores than during other phases of the moon 

 (Hobson et al. 1981). The number of grunts settling 

 during full moons should be low, which it is (Fig. 4). 

 In contrast, if length of darkness is a dominant factor 

 to successful settlement, then the highest influx of 

 recruits should coincide with the new moons, and 

 intermediate numbers with quarter moons, which 

 they do not (Fig. 4). The recruitment dynamics, 

 nevertheless, indicate higher influxes of PL-l's when 

 some degree of darkness occurs during each night 

 (Fig. 4). We hypothesize that recruitment is depen- 

 dent primarily on favorable currents, with predation 

 possibly acting as a secondary selective force. 



Reversals in surface currents and oppositely flow- 

 ing currents in the upper 30 m, which could affect 

 the dispersal of larvae, are known to occur at St. 

 Croix and at Puerto Rico (Gladfelter et al. 1978; Lee 

 et al. 1978; Molinari et al. 1980). Their local patterns, 

 however, remain unknown. Eddy formation to the 

 west of the island of Barbados, produced by Karmen 

 trails as the generally west-setting current passes 

 the island, has also been postulated to retain the pela- 

 gic stages of inshore species (Emery 1972). It is 

 reasonably well established that seasonal shifts in 

 local current gyres in the vicinity of Hawaii favor the 

 settlement of various species of reef fishes (Sale 

 1970, 1980; Johannes 1978). In addition, spawning is 

 often synchronized to disperse eggs and lan-ae away 

 from reefs and into offshore currents (Johannes 

 1978; Lobel 1978). But specific currents do not 

 always trigger spawning, as Colin (1982) reported 

 for several reef fishes at Puerto Rico. In these in- 

 stances, rhythmic spawning often can be related to 

 the lunar cycle, but significantly, some species spawn 

 every day. Depending on the time of spawning, 

 dispersal routes for fish eggs and larvae can vary 

 over short-time periods because of reversals in cur- 

 rents. 



The daily behaviors and distributions of grunts at 

 sea are unknown. Do larval grunts passively drift 

 with currents? Or do they seek different depths at 

 different times of the day? Active behaviors that 

 would utilize differences in currents have been in- 

 voked to explain the retention of pelagic larval fishes 

 and invertebrates close to the island of Oahu (Leis 

 1982). Similar activities by larval French grunts 



423 



