Eggleston et al : Artificial shelters and survival of juvenile Panulirus argus 



693 



19°50' 



habitats were chosen to compare 

 relative rates of predation: an 

 inner-bay sand-seagrass (Thalas- 

 sia tesUidinum) flat located at 

 the northwestern portion of the 

 bay, and an outer-bay seagrass 

 meadow adjacent to a coral reef 

 (Fig. 2). Seagrass and algal habi- 

 tats likely provide the only natu- 

 ral daytime refuge for juvenile P. 

 argus in this system because of 

 an apparent lack of crevices 

 (formed by rocky outcrops, patch 

 coral reefs, sponges, solution 

 holes, or undercut seagrass 

 banks). Anecdotal information 

 from lobster fishermen present 

 in Bahia de la Ascension prior to 

 the introduction of casitas 

 (around 1974) indicated that ju- 

 venile lobsters commonly resided 

 solitarily under dense stands of 

 Thalassia or complex red algae 

 (e.g., Laurencia), or aggregated 

 around existing structures such 

 as sponges or cobble. Moreover, 

 previous tethering experiments 

 with juvenile P. argus in this 

 system demonstrated that sea- 

 grass and algae provide some 

 protection for spiny lobster 

 juveniles from predators (R.N. 

 Lipcius et al., unpubl. data). 



Differences in seagrass density between and within 

 sites were determined prior to experiments by measur- 

 ing dry-weight biomass (g) of Thalassia removed from 

 0.25 m^ plots. The inner-bay site was composed of 

 sparse seagrass patches (x Thalassia biomass 62.4 

 g/m-. A'' 6, SD 10.7) interspersed among coarse cal- 

 careous sand and coral rubble. The coral rubble was 

 covered mostly by green and red algae (Dasycladus 

 spp. and Laurencia spp., respectively), but also sup- 

 ported larger sponges. The outer-bay site was located 

 shoreward of a fringing coral reef and composed of 

 sand patches and patch corals interspersed among 

 moderate to dense seagrass beds (i Thalassia bio- 

 mass 111.6g/m2, N 6, SD 13.4, and 210.0g/m2, N 6, 

 SD 12.6, respectively). Further details of the study site 

 are described in Eggleston et al. (1990). 



Artificial shelters 



Our design of artificial lobster shelters was based on 

 "casitas Cubanas"— sunken wood and concrete struc- 

 tures that simulate lobster dens (Miller 1989) (Fig. 1). 



87°30' 



-" KReef 



GULF OF '\_J 

 MEXICO 



CUBA 



€> 



Figure 2 



Study sites at Bahia de la Ascension, Mexico. 



The large casitas used in this study (177x118x6cm) 

 were constructed with a reinforced concrete roof bolted 

 to a supporting PVC-pipe frame. Several physical prop- 

 erties of the casita appear to make it an optimal lobster 

 den: (1) shaded cover provided by the wide concrete 

 roof, (2) a low ceiling that excludes large piscine 

 predators, and (3) multiple den openings which are 

 smaller than the inner roof height of the casita (Fig. 

 1) (Eggleston et al. 1990). Hence, the use of casitas per- 

 mitted us to standardize den size and availability in dif- 

 ferent habitats. 



Tethering experiments 

 and predator observations 



Spiny, lobsters were collected from existing casitas and 

 held in traps for 1-2 days prior to initiation of each ex- 

 periment. Only intermolt lobsters exhibiting strong 

 "tail flipping" responses were used in tethering experi- 

 ments. Tethers were constructed by locking a plastic 

 cable-tie around the cephalothorax of a lobster, be- 

 tween the second and third walking legs, and securing 

 the cable-tie with cyanoacrylate cement. The cyano- 



