654 



Fisherv Bulletin 100(4) 



malin. Morphometric and meristic characters were exam- 

 ined as described in Smith (1971) by using dial caHpers 

 and a meter stick. The morphometric data were analyzed 

 by using a sheared principal component analysis with a 

 covariance matrix to confine the effect of size to the first 

 principal component (Humphries et al., 1981; Bookstein 

 et al., 1985; Stauffer et al., 1997). The meristic data were 

 analyzed by using a principal component analysis with a 

 correlation matrix. 



Allozyme analysis 



Horizontal starch gel electrophoresis followed the pro- 

 tocols described in Murphy et al. (1996). Gels (12'7fw/v; 

 Starch Art Corp., Smithville, TX) were run on one of three 

 buffer systems: Tris-citrate II buffer (TC II) (30 niAmps for 

 14 hours), lithium hydroxide buffer ( LIOH) (25 niAmps for 

 14 hours), or Tris borate-EDTA buffer (EBT) (30 mAmps 

 for 14 hours). Histochemical staining followed the proto- 

 cols of Murphy et al. (1996), and locus nomenclature and 

 allelic designations followed Shaklee et al. ( 1990). 



A preliminary survey of 16 loci in 15 individuals each of 

 C. fidva and P. furcifer (Table 2) was performed to identify 

 those loci for which the alleles were consistently differ- 

 ent among the presumed parent species, C. fulva and P. 

 furcifer. All parent individuals and putative hybrids were 

 then surveyed for all loci that demonstrated differences 

 between the species. 



Nuclear DNA analysis 



An actin gene intron and the second intron in the S7 

 ribosomal protein gene (Chow and Hazama, 1998) were 

 investigated by using restriction fragment length poly- 

 morphism (RFLP) analysis. Amplification primers and 

 reaction conditions are listed in Table 3. The regions were 

 amplified by using the PCR reagent system (GIBCO/BRL 

 Life Technologies®, Bethesda, MD) and a 25-pL reaction 

 cocktail ( IX PCR buffer with MgCl.^, 0.2 mM dNTP, 0.5 

 pM primer, 2.5 U of Tag DNA polymerase, and 25-50 ng 

 genomic DNA template). Some PCR reactions were per- 

 formed with Platinum® Tag high fidelity (GIBCO/BRL 

 Life Technologies®) with a 25-pL reaction cocktail (IX 

 high fidelity buffer, 2 mM MgSO_,, 0.2 mM dNTR 0.2 pM 

 primer, and 2.5 U of Platinum® Tag DNA polymerase high 

 fidelity). In other cases, 1 pL dimethyl sulfoxide (DMSO, 

 Fisher Scientific BP231-1, Pittsburgh, PA) was added to 

 the reaction to increase sensitivity 



The amplification products from two individuals of each 

 species for both loci were digested with a panel of restriction 

 enzymes to identify those that exhibited differences between 

 the putative parent species. All samples were subsequently 

 digested with those enzymes that demonstrated differences 

 in the pilot study (Table 3). Digestion reactions ( 1.5-pL lOX 

 buffer, 3 U restriction enzyme, and 4-pL PCR product) were 

 incubated 2 to 18 hours at 37°C. Digestion products were 

 separated on 2.5*7^ agarose gels (1.25% Ultrapure Agarose, 

 GIBCO/BRL Life Technology I R) + 1.25% NuSieve GTG (R) 

 agarose, FMC Biochemical, Rockland, ME), stained with 

 ethidium bromide and visualized under UV light. 



MtDNA analysis 



The following regions of the mitochondrial genome were 

 surveyed in the three putative parent species and hybrids; 

 the adenosine 5'-triphosphatase subunit 6 (ATPase 6) 

 gene, the 12S/16S ribosomal RNA gene region, and the 

 nicotinamide dehydrogenase subunit 4 (ND4) gene. Ampli- 

 fication primers and reaction conditions are listed in Table 

 3. Amplified products of two individuals of each putative 

 parent species were screened with restriction enzymes to 

 identify potential differences between C. fulva and P. fur- 

 cifer (Table 3). All individuals were screened at the three 

 regions with those enzymes that revealed differences in 

 the preliminary study. Restriction digestion reactions were 

 performed and visualized as described for nuclear DNA. 

 For each individual, the haplotype designations of each 

 region were combined in sequence creating a composite 

 haplotype. 



Molecular data analysis 



Nei's (1978) unbiased genetic distance was calculated 

 from the allozyme data by using the computer program 

 BI0SYS2 (Swofford and Selander, 1989). Mean nucleotide 



