Laserfiche WebLink
relatively young as determined by ototlith analysis (Minckley, unpublished data), and thus <br />may reflect conditions following the last developments in the system (e.g., Flaming Gorge <br />and Glen Canyon dams). <br />The decline in variation within populations as one moves upstream from Lake Mohave is <br />explainable by at least two hypotheses. The Lake Mohave region could represent a "sink" <br />where variation generated in several populations of upstream regions (e.g., Green, Yampa, and <br />upper Colorado rivers) may have accumulated in Lake Mohave through downstream dispersal <br />(Figure 3A). This hypothesis is, however, unlikely for several reasons. Production of the <br />diversity in Lake Mohave would require that each upstream tributary contains one or several <br />populations, each with their own unique suite of haplotypes. Under this model, therefore, <br />populations from the Green/Yampa should be diagnosably distinct from those of the upper <br />Colorado River, a situation which does not exist. In addition, the observed increase in <br />diversity between Lakes Powell and Mohave would require existence of drainages entering <br />the Colorado River below Lake Powell, each with a unique set of haplotypes. The only two <br />tributaries of any appreciable size aze the Little Colorado and Virgin rivers, for which no <br />records of large populations of razorback suckers exist. <br />The most significant objection to this hypothesis, however, results from consideration of <br />past climatic conditions. The last full glacial period peaked 18 thousand years ago (Ka) and <br />ended approximately 10 Ka (Winograd et al., 1992). Warm season (April-September) air <br />temperatures on the Colorado Plateau at that time averaged 6.3°C cooler than today <br />(Betancourt, 1990). Although difficult to translate into water temperatures, especially since <br />discharges are inestimable (except that they were higher), it follows that river temperatures <br />would have also had lower averages and extremes. Spawning, hatching, and "optimal" <br />growth of razorbacks and other "big-river" fishes show negative effects under laboratory and <br />12 <br />