700 



Fishery Bulletin 90(4). 1992 



on small juvenile lobsters tethered in seagrass suggests 

 that sparse-to-moderate-density Thalassia does not 

 provide adequate protection from predators, and that 

 the addition of shelter greatly enhances survival for 

 these smaller juvenile lobsters. Thus, the use of arti- 

 ficial lobster shelters in sparse-to-moderate-density 

 Thalassia beds may effectively reduce predation- 

 induced mortality rates of small juvenile lobsters and 

 thereby enhance production of this size-class. However, 

 given the general relationship of increasing survival 

 with habitat complexity for many decapod crustaceans 

 (Heck and Thoman 1981, Wilson et al. 1987, Heck and 

 Crowder 1991 and references therein), the relative im- 

 portance of shelter availability upon survival of small 

 juvenile lobsters may be reduced in habitats with dense 

 Thalassia. Thus, further studies are required to under- 

 stand the relationship between shelter availability and 

 increasing habitat complexity upon survival of small 

 juvenile lobsters. 



The reduced survival of large juvenile lobsters near 

 casitas compared with seagrass 60-70 m away during 

 the January 1989 experiment is consistent with our 

 previous results for this lobster size-class. For ex- 

 ample, survival of small lobsters (46-55 mm CL) in large 

 casitas was significantly higher than survival of large 

 lobsters (56-65 mmCL) (Eggleston et al. 1990). More- 

 over, large lobsters survived better in medium than in 

 large casitas (Eggleston et al. 1990). Eggleston et al. 

 (1990) suggested that medium casitas excluded pred- 

 ators that were able to prey on large lobsters, and 

 postulated that larger predators associated with large 

 casitas may selectively prey upon larger lobsters, due 

 to better visual perception with increasing predator 

 and prey size (Kao et al. 1985, Ryer 1988). The signifi- 

 cant positive correlation between the numbers of 

 predators (primarily gray snapper L. griseus) occupy- 

 ing specific casita stations and predation rates at these 

 same stations suggests that gray snapper may be the 

 principle predator of juvenile lobsters inhabiting casitas 

 at the inner-bay nursery site. Gray snapper (15cmTL) 

 have successfully attacked small early-juvenile lob- 

 sters tethered in Florida Bay (Herrnkind and Butler 

 1986). 



The combined results from this study and previous 

 work in Bahia de la Ascension, Mexico (Eggleston et 

 al. 1990), suggest that juvenile lobsters would survive 

 better by leaving large shelters to take up residence 

 in smaller shelters or nearby seagrass habitats when 

 they reach a body size of ~56-65mmCL. This idea of 

 enhancing survival through size-specific emigration 

 from large shelters was partially supported during our 

 recent observations of habitat-specific and size-specific 

 patterns of shelter use by juvenile P. argus in Bahia 

 de la Ascension, Mexico. Our recent field observations 

 (Eggleston and Lipcius 1992) indicated that shelter- 



seeking behavior of P. argus is highly flexible to local 

 social conditions (i.e., presence of conspecifics) and 

 shelter scaling. For example, in a habitat containing 

 very few conspecifics (e.g., outer-bay site), large juve- 

 nile lobsters chose smaller, safer medium casitas over 

 large casitas as predicted by our tethering results (this 

 study; Eggleston et al. 1990). However, in a habitat 

 containing large numbers of conspecifics (e.g., inner- 

 bay site), large juvenile lobsters occupied large casitas 

 with large conspecifics (Eggleston and Lipcius 1992). 

 The tethering technique in this study did not address 

 the potential benefits of gregarious residency to lobster 

 survival. Gregarious occupancy by more than the six 

 tethered lobsters appeared to be inhibited because of 

 the tethering technique, i.e., lobsters did not colonize 

 casitas containing tethered individuals (pers. observ.). 

 Since gregarious sheltering has been implicated as a 

 mechanism for reducing predator-induced mortality 

 (Berrill 1975, Herrnkind et al. 1975, Eggleston and Lip- 

 cius 1992), final conclusions regarding the impact of 

 casitas upon predation-induced mortality rates of large 

 juvenile lobsters must not only consider the size-specific 

 relationship between shelter-associated predators and 

 lobsters, but also the potential benefits of gregarious 

 sheltering. 



Results from our August 1989 experiment support 

 the hypothesis that the impact of artificial shelters upon 

 predation-induced mortality of juvenile lobsters varies 

 according to the distance of unprotected lobsters from 

 these shelters. During the August experiment at the 

 outer-bay site, small lobsters survived equally well 

 whether they were tethered beneath casitas or 30 m 

 away. These tethering results, combined with obser- 

 vations on predator movements, suggest that 30 m is 

 beyond the daytime foraging range of most casita- 

 associated predators. However, the lack of a signifi- 

 cant correlation between the numbers of potential 

 predators at a specific casita station and predation 

 rates on lobsters at these same stations at the outer- 

 bay site during the January 1989 experiment suggests 

 that transient predators such as jacks (Caranx spp.), 

 groupers (Epinephelus spp.), sharks (Ginglymostoma 

 cirratum. and Sphyrma spp.), and stingrays (Dasyatis 

 spp.) may be moving from the nearby barrier reef (see 

 Figs. 3 and 4 for geography) and preying on tethered 

 lobsters. Gut contents of stingrays (Dasyatis spp.) and 

 bonnethead sharks S;)/i,i/rwa tiburo, captured at night 

 in nearshore Florida Bay waters, contained a high pro- 

 portion of early-juvenile spiny lobsters (Smith and 

 Herrnkind 1992). Nurse sharks Ginglymostoma cir- 

 ratum are also known predators of juvenile P. argus 

 (Cruz and Brito 1986). Thus, our observations on the 

 daytime abundance and movements of casita-associated 

 predators (i.e., primarily mutton and yellowtail snap- 

 per, Lutjanus analis and 0. chrysurus) at the outer- 



