256 Conservation Biology, Pages 256–259 Volume 11, No. 1, February 1997 ‡ Address correspondence to this author at his current address: De- partment of Biology, Denison University, Granville, OH 43023, U.S.A., email ramirez@denison.edu § Current address: Department of Integrative Biology, University of California, Berkeley, CA 94720, U.S.A. Paper submitted September 22, 1995; revised manuscript accepted April 10, 1996. Minimal Genetic Variation in a Coastal Dune Arthropod: The Trapdoor Spider Aptostichus simus (Cyrtaucheniidae) MARTIN G. RAMIREZ*‡ AND JULIE L. FROEHLIG†§ *Department of Biology, Pomona College, Claremont, CA 91711, U.S.A †Division of Math and Science, Walnut High School, Walnut, CA 91789, U.S.A. Introduction Coastal dunes originally occurred at disjunct sites along 23% of California’s 1326-km coastline (Cooper 1967), but human disturbance has severely reduced or elimi- nated many dune systems (Powell 1981). Indeed, coastal dunes are considered to be among the most rapidly dis- appearing habitat types in California (Schoenherr 1990). Thus, whereas California coastal dunes have always been “island” habitats in a “sea” of land, they have be- come increasingly fragmented and insular as a result of human activities. Knowledge of the distribution of ge- netic variation within and among populations of organ- isms residing in such remnant coastal dunes will be nec- essary if we are to assure their long-term existence. The trapdoor spider Aptostichus simus inhabits coastal dunes of southern California and the California Channel Islands (Ramirez 1995). It lives in burrows con- centrated in and about stands of native dune vegetation and extending into the dunes amid litter and the root systems of the plants. We present the results of an analy- sis of the amount and distribution of genetic variation in this coastal dune endemic. Methods During June-July 1992 we collected A. simus from nine sites in southern California (Fig. 1). At each site A. simus were collected by sifting sand below patches of dune vegetation. We collected 15 spiders from each popula- tion for a total of 135. In the laboratory spiders were starved for a week and then frozen at -75° C until they were prepared for electrophoresis. Enzyme electrophoresis generally followed Ramirez (1990). Gels were 12.5% starch. Twelve enzymes en- coded by 13 genetic loci were analyzed: adenylate kinase (ADKIN), alcohol dehydrogenase (ADH), fumarase (FUM), glucosephosphate isomerase (GPI), glyceraldehyde-3- phosphate dehydrogenase (G-3-PDH), hexokinase (HK), isocitrate dehydrogenase (IDH), malate dehydrogenase- 1,2 (MDH-1,2), mannose-6-phosphate isomerase (MPI), peptidase (PEP), phosphoglucomutase (PGM), and su- peroxide dismutase (SOD). These enzymes were resolved using three buffer systems: Discontinuous Tris-Citrate (Poulik 1957) [ADH, HK, PEP, PGM, SOD]; Continuous Tris-Citrate I (Selander et al. 1971) [GPI, G-3-PDH]; and Tris-Maleate (Selander et al. 1971) [ADKIN, FUM, IDH, MDH, MPI]. We used the BIOSYS-1 computer package (Swofford & Selander 1981) computer package to analyze the electro- phoretic data. Agreement between population geno- typic proportions and Hardy-Weinberg expectations was evaluated by chi-square tests for goodness of fit. Genetic identity (I) coefficients (Rogers 1972) between each pair of populations were calculated and used as a measure of population differentiation. The apportionment of ge- netic differentiation among populations was analyzed by use of Wright’s (1965) F ST statistic as modified by Nei (1977). We calculated all F ST values using means and variances of allele frequencies weighted by sample sizes. We estimated gene flow (Nm) from the F ST values, using the equation Nm = (1 - F ST )/4F ST (Wright 1951). Results Eleven loci (ADH, ADKIN, FUM, G-3-PDH, HK, IDH, MDH-1, MPI, PEP, PGM, SOD) were monomorphic