Microhabitat associations of three sympatric species of Naucoridae (Insecta: Hemiptera)

Samples (0.5 m2) were taken from the South Llano River, Kimble County, Tex., to determine inter- and intraspecific differences in microhabitat association for the three most abundant species of Naucoridae. Various aspects of substratum condition, current speed, temperature, depth, as well as complex key hydraulic features, were used to characterize the microhabitat for each sample. Eight species of naucorids occur in the river; however, only Ambrysus circumcinctus Montandon, Cryphocricos hungerfordi Usinger, and Limnocoris lutzi La Rivers occurred with sufficient frequency in our samples to allow statistical analysis. None of the three species exhibited age-related differences in microhabitat association based on multivariate analysis of variance (MANOVA), and none of the examined abiotic variables yielded significant differences between nymphs and adults when analyzed from a univariate perspective except Froude number for L. lutzi. Both A. circumcinctus and C. hungerfordi exhibited highly significant differences (MANOVA) in abiotic characteristics between quadrats of occurrence and nonoccurrence. Results of individual analyses of variance (ANOVAs) indicated that 9 of the 10 abiotic variables were significant for A. circumcinctus and 8 for C. hungerfordi. In contrast, quadrats where L. lutzi occurred could not be distinguished from those in which it was absent based on MANOVA, and only two abiotic variables exhibited significance in the ANOVAs. Associations between age groups (nymphs and adults) within a species and between all possible species pairs (combined age groups) also were evaluated with contingency X2 tests using presence-absence data and with Pearson's product-moment correlations using density data. In all cases, significant negative associations were not detected, suggesting that neither age groups nor species exhibit microhabitat segregation. Discriminant function analysis for each species supported MANOVA results on presence-absence data. Separate multiple step-up regressions were used to evaluate which abiotic factors were most related to the density for each species. Significant variables included mean current speed, which accounted for 23.8% of the variation in density of A. circumcinctus; rock standard error of the mean, which accounted for 43.1% of the variation in density of C. hungerfordi; Froude number and depth range, which together accounted for 39.0% of the variation in density of L. lutzi. In summary, each common naucorid clearly shows microhabitat associations, but the evidence of interspecific differences is indirect.