NOTE San/er et al.: Occurrence of Panulirus argus westonii in Florida waters 



871 



labeled primers. All sequences were deter- 

 mined in both directions. 



Aligned DNA sequence data included 

 441 sites. Distance and parsimony anal- 

 yses of the aligned DNA sequences were 

 done with PAUP (Swofford, 1998). Three 

 separate distance calculations were made: 

 uncorrected mean. Jukes-Cantor (Jukes 

 and Cantor, 1969), and Kimura 2-param- 

 eter (Kimura, 1980) distances. Neighbor- 

 joining trees (Saitou and Nei, 1987) were 

 generated by using all three distance mea- 

 sures. Parsimony trees were produced by 

 using branch and bound searches. Sites 

 coded as gaps were treated either as miss- 

 ing, or as a fifth character state. The 

 mitochondrial 16S rDNA sequences were 

 aligned by using the SeqPup DNA analy- 

 sis program (Gilbert-). Support for nodes 

 of both parsimony and distance trees was 

 assessed by calculating bootstrap propor- 

 tion (BP) values (Felsenstein, 1985) from 

 1000 replicate searches by using the same 

 tree building methods used to produce 

 the trees. Scyllarides nodifer (Stimpson) 

 (Decapoda, Scyllaridae) was used as an 

 outgroup for phylogenetic analyses. 



Results 



100 



92 



Pargwest-2 



97 



90 



Pargwest-5 



Pargwest-4 



Pargwest-1 



' Pargwest-3 



Mia226 

 V22 

 >- Mia204 

 r Pargarg-1 



' Pargarg-2 



'■ Pargarg-5 

 ■— Pargarg-4 

 — Pargarg-3 



FLORIDA 



VENEZUELA 

 FLORIDA 



Panulirus argus 

 spp westonii 



Pargarg-6 



Panulirus argus 

 spp. argus 



Scyllarides 

 nodifer 



0,02 



Figure 1 



Neighbor joining tree derived from the Kimura two-parameter distances cal- 

 culated from partial mitochondrial 16S rDNA data for 14 Panulirus argus 

 and Scyllarides nodifer samples. Bootstrap values calculated from the dis- 

 tance analysis are shown above and bootstrap values calculated from parsi- 

 mony analysis are shown below strongly supported branches. 



The mitochondrial 16S rDNA region ana- 

 lyzed in our study included 441 aligned nucleotide posi- 

 tions. In addition to single base differences, the complete 

 aligned data set included 19 presumptive insertion or dele- 

 tion mutations (indels) of 1-5 bases in length. The majority 

 of indels involved only one base ( 13 of 19) and only six were 

 potentially informative. 



Regardless of the type of distance calculation, there was 

 little variation in the outcome of the different distance cal- 

 culations (less than 1%). Results from an analysis with a 

 PCimura 2-parameter model yielded distance values rang- 

 ing from 5.9'7f to 15.1% between the Caribbean and Bra- 

 zilian subspecies of Panulirus argus (P. argus argus and 

 P. argus westonii respectively). For the three genetically 

 distinct Caribbean P. argus from Silberman et al. ( 1994a), 

 sequence divergence estimates indicated closer affinities 

 to Brazilian P. argus (2.6% ) than to the Caribbean P. argus 

 (7.4%). 



Phylogenetic analysis of Brazilian, Caribbean and the 

 three divergent Caribbean P. argus indicated a tree topol- 

 ogy that places the three divergent Caribbean samples 

 off! argus in the Brazilian clade (Fig. 1). This basic tree 

 topology is strongly supported regardless of the method of 

 phylogenetic analysis. 



Gilbert, D. G. 1996. SeqPup biosequence editor and analysis 

 application, vers. O.Gf Department of Biology, Univ. Indiana, 

 Bloomington, IN 47405. 



Discussion 



Traditionally Panulirus argus has been thought to be dis- 

 tributed from North Carolina in the western Atlantic, 

 throughout the Caribbean to Rio de Janeiro, Brazil. Sarver 

 et al. (1998) presented data that questioned the taxo- 

 nomic status of P. argus. DNA sequence data and mor- 

 phological differences suggest that populations of P. argus 

 from Brazil are genetically distinct from Caribbean popu- 

 lations and that the level of divergence is equivalent to 

 the levels of divergence seen between recognized species of 

 Panulirus. As a result they suggested formally recognizing 

 two genetic forms of P argus and have recommended pro- 

 visional subspecific status: P. argus argus (Caribbean) and 

 P. argus westonii (Brazil) until a formal taxonomic revision 

 can be done. The occurrence of cryptic species in Panulirus 

 is not restricted to P. argus. A similar situation exists in P. 

 longipes, where there appears to be at least four recogniz- 

 able forms of uncertain taxonomic status (George, 1972; 

 Sekiguchi, 1991; Chan and Chu. 1996). Panulirus homa- 

 rus has also been broken into three geographic forms that 

 are given the rank of subspecies (Berry, 1974). 



We can only speculate on the events that led to this dis- 

 tribution of the two P. argus subspecies. During the late 

 Miocene Epoch, a P. argus ancestor may have occupied the 

 North Atlantic and as Africa and South America moved 

 apart as a result of continental drift, habitat became avail- 

 able for P. argus along the South America coast. These pop- 



