Laserfiche WebLink
<br />I <br />I <br />I <br />I <br />I I <br />I !if <br />t <br /> .. <br />I .. <br />" <br />I ~ <br />... <br />'0 <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br /> - <br /> <br /> <br />40 <br /> <br />and examined directly with transmitted light at 160 to 1000 X. The structures are small <br /> <br />and transparent enough that grinding or polishing was unnecessary. Longer focal lengths <br /> <br />at lower magnifications typically provided better resolution through curved otolith <br /> <br />surfaces; higher magnification was required to resolve questions of increment separation. <br /> <br />Clearing through penetration of immersion oil increased contrast within a few minutes <br /> <br />and enhanced analysis for a number of hoW'S. Increments, defined by contrasting light <br /> <br />and dark regions, were counted at least twice for each otolith. Otoliths "over-cleared" in <br /> <br />immersion oil within a day, with increments becoming difficult to count as the medium <br /> <br />and otolith became more similar in light refraction. A plastic mounting medium suffered <br /> <br />the same drawback and was deemed unusable. <br /> <br />Sagittae, at least for younger larvae, proved most usable because of their rapid <br /> <br />increase in size. Lapilli counts were only made when questions of increment clarity <br /> <br />resUlted in problems. For unknown reasons, accurate numbers of increments could not be <br /> <br />ascertained in 10 to 15% of all otoliths examined. <br /> <br />Time of first increment formation was estimated by least-squares regression of <br /> <br />total increments with known age of larvae in days as the dependent variable. The Y-axis <br /> <br />intercept was the time of first increment deposition. Rate of increment formation was the <br /> <br />slope of the regression; a slope significantly different from 1.0 indicated other than daily <br /> <br />formation. <br /> <br />Determination of increment deposition based on growth rate was done by com- <br /> <br />parison of slopes and intercepts of regressions through use of analysis of covariance <br /> <br />(ANCOV A) across the three feeding treatments. The ANCOV A interaction term was <br />