MORPHOLOGICALLY CRYPTIC SPECIES CONFOUND ECOLOGICAL-STUDIES OF THE CADDISFLY GENUS GUMAGA (TRICHOPTERA, SERICOSTOMATIDAE) IN NORTHERN CALIFORNIA

Two species in the caddisfly genus Gumaga (Sericostomatidae) are curre ntly recognized in North America: Gumaga nigricula (McLachlan, 1871) a nd Gumaga griseola (McLachlan, 1871). Ecological and morphological stu dies over the last 20 years indicated that both species were either ex tremely variable or that morphologically cryptic species were unknowin gly being included in the studies. The study reported here examined wh ether genetic characters las measures of reproductive isolation and ge netic differentiation) could resolve the taxonomic issues concerning G umaga, and consequently provide insight into the observed ecological a nd morphological variation. Allozyme electrophoresis was used to exami ne the genetic relationships among larvae of Gumaga collected from Ave streams and two springbrooks in northern California. For each specime n. 18 enzymes representing 21 presumptive,oene loci were scored. Genet ic variability was high at all but one site: 14.3-47.6% of the loci we re polymorphic (3-10 loci per site) and heterozygosity averaged 5.9-20 .7%. Six genetically distinct groups of individuals were identified (i .e., Gumaga types A. B, C, D, E, F). Mean Nei's genetic distances betw een groups ranged from 0.371 (type A versus type B) to >1.0 (type F ve rsus types A, B, D, or E). The high degree of genetic differentiation among groups is maintained even when the groups are in close proximity (e.g., Gumaga types A and E at the same site and Gumaga types A, B, a nd C within the same drainage basin). In addition, previous studies ha ve found evidence of premating mechanisms that limit interbreeding amo ng Gumaga types A, B, and D. Thus, it appears that these six groups re present reproductively isolated species rather than genetic variants o f one or two species. Extensive morphological and genetic studies are necessary to clarify taxonomic relationships within the genus Gumaga, but the results of this and other genetic analyses of aquatic insects illustrate the potential insight that this approach can provide to tax onomic, behavioral, and ecological studies. Furthermore, these results also illustrate how unusual land what is often perceived as interesti ng) ecological variability observed for a single species may in fact r eflect the presence of morphologically cryptic species.