Laserfiche WebLink
SENSITIVITY ANALYSiS~ <br />To determine- how sensitive the model is to average survivorship values and the <br />variation (standard deviation) around them for each age class, we decreased <br />survivorship by 5% and 10% and increased standard deviation by 10% and 20% for <br />each age class separately. With each parameter, we ran 100 simulations and <br />tabulated how many times the three squawr'ish subpopulations (Upper Green, Lower <br />Green and Colorado) fell below 125, 250 and 500 total adults (tables 5-7). These <br />simulations provide both a measure of how sensitive the model output is to small <br />changes in the input parameters and also a probability estimate regarding the <br />persistence of the population. We chose 125, 250 and 500 adults as the cut off value <br />because these numbers represent 25, 50 and 100 adult females (recall that the <br />estimated sex ratio is 5 males: 1 female}. These numbers represent minimum numbers <br />of females used by conservation biologists for recovery programs. Decreasing the <br />survivorship values for all age groups resulted in approximately .33 of the simulations <br />for the Colorado River squawfish subpopulations to decrease below 125 individuals <br />(Table 5). The Green River subpopulations did not drop below 125 individuals except <br />when adult survivorship was decreased. Decreasing survivorship by 5 or 10% resulted <br />in a high probability of both the Colorado River and Lower Green subpopulations falling <br />below 500 adults (100 breeding females). The model is not as sensitive to changes in <br />annual variation of survivorship values, although increases of 10 or 20°~ in variance did <br />result in approximately 90% of the simulations resulting in adult populations of the <br />21 <br />