Laserfiche WebLink
31 <br />selective feeding by larval razorback suckers, following Hunter <br />(1981). Biomass of invertebrates was expressed as mm3/m3. Values for <br />numbers and biomass were weighted by habitat volumes to derive means <br />for entire ponds. <br />Quantitative data on dispersion of larvae were obtained from <br />plarkton tom,, but to ensure adequate sample sizes of fish, a bright <br />spotlight was used at night to attract greater numbers of the <br />phctotactic larvae (Marsh and Ianghorst 1988). During weeks 1 and 2, <br />both day and night samples of fish were analysed to obtain close-order <br />data on initial feeding (week 1) and adequate sample sizes (week 2). <br />Night collecting of fish was perfected by week 3. Daytime collecting <br />was by then ineffective and was terminated. <br />Larvae were preserved in 51 buffered formalin solution. Upper lip <br />length (from tip of upper lip to lateral junction of upper and lower <br />lips), used as an index of mouth size (Shirota 1970), and TL, were <br />measured by ocular micrometer at aPPropriate magnification on a <br />dissecting microscope. Preserved specimens were blotted dry and <br />weighed on a Mettle H-57 balance to the nearest 0.01 mg. Guts were <br />excised and all food items were identified and enumerated. As in <br />plankton samples, invertebrates were measured (width and length), and <br />assigned a gemietric shape to reconstruct volume. Strauss' (1979, <br />1982) selectivity index (Li) was used to evaluate potential selection <br />for species and size classes of food items. <br />Ponds were drained and fish recovered 18-19 April 1985. Larval <br />tiger salamanders (Ambystoana tigrinum), originating from natural