Academia.eduAcademia.edu

Figure 5 – uploaded by Frank Rahel

See full PDFdownloadDownload figure
Ficure 5.—Isopleths of simulated power (1 — 8) to detect annual declines in the biomass of trout in streams of the Medicine Bow National Forest, Wyoming, when the number of sites and number of years monitored is varied. The mean temporal variation (CV) used in the simulations was 54%. The comparisons across rows show the effect of reducing the annual rate of decline from 5.0% to 2.5%; the comparisons down columns show the effect of increasing the type I error rate («) from 0.05 to 0.10 and 0.20.

Figure 5 Ficure 5.—Isopleths of simulated power (1 — 8) to detect annual declines in the biomass of trout in streams of the Medicine Bow National Forest, Wyoming, when the number of sites and number of years monitored is varied. The mean temporal variation (CV) used in the simulations was 54%. The comparisons across rows show the effect of reducing the annual rate of decline from 5.0% to 2.5%; the comparisons down columns show the effect of increasing the type I error rate («) from 0.05 to 0.10 and 0.20.

Related Figures (6)

FicureE 1.—Steps in simulating the statistical power of a monitoring program to detect trends in trout abundance at a single site or a network of multiple sites. In step 1, a declining trend was projected onto an initial abundance estimate for a site for the projected time frame of monitoring. For each year, open circles represent the mean abundance and lines represent one standard deviation. In step 2, random variation was added to the abundance estimate for each year (closed circles) by using a random deviate from a normal distribution with the mean equal to the projected abundance and the standard deviation approximated for the initial abundance estimate. In step 3, the slope of the least-squares regression line was estimated for each site. For single sites, we determined whether the slope was different from zero. For multiple sites, we determined whether the arithmetic mean slope among sites was different from zero. We repeated each step 1,000 times; the proportion of times a slope was different from zero represents the statistical power and assesses how often known trends are detected amid simulated variation. The figure was adapted from Gibbs (1998).
FicureE 1.—Steps in simulating the statistical power of a monitoring program to detect trends in trout abundance at a single site or a network of multiple sites. In step 1, a declining trend was projected onto an initial abundance estimate for a site for the projected time frame of monitoring. For each year, open circles represent the mean abundance and lines represent one standard deviation. In step 2, random variation was added to the abundance estimate for each year (closed circles) by using a random deviate from a normal distribution with the mean equal to the projected abundance and the standard deviation approximated for the initial abundance estimate. In step 3, the slope of the least-squares regression line was estimated for each site. For single sites, we determined whether the slope was different from zero. For multiple sites, we determined whether the arithmetic mean slope among sites was different from zero. We repeated each step 1,000 times; the proportion of times a slope was different from zero represents the statistical power and assesses how often known trends are detected amid simulated variation. The figure was adapted from Gibbs (1998).
TasLe 1.—Coefficients of variation (CVs) for the abundance and biomass of trout species over time. Data for all species combined are presented when more than one species was reported per stream.
TasLe 1.—Coefficients of variation (CVs) for the abundance and biomass of trout species over time. Data for all species combined are presented when more than one species was reported per stream.
TABLE 1.—Continued.
TABLE 1.—Continued.
Figure 3.—Simulated power (1 — B) to detect annual declines of 2.5% and 5% in the abundance or biomass of a trout population after 5-25 years, given the temporal variation in the population parameter represented by the coefficient of variation and the type I error rate («). The comparisons across rows show the effect of decreasing the magnitude of the annual decline from 5% to 2.5%; the comparisons down columns show the effect of increasing « from 0.05 to 0.10 and 0.20.
Figure 3.—Simulated power (1 — B) to detect annual declines of 2.5% and 5% in the abundance or biomass of a trout population after 5-25 years, given the temporal variation in the population parameter represented by the coefficient of variation and the type I error rate («). The comparisons across rows show the effect of decreasing the magnitude of the annual decline from 5% to 2.5%; the comparisons down columns show the effect of increasing « from 0.05 to 0.10 and 0.20.
Related topics:
Social PsychologyBiologyPopulation TrendPopulation MonitoringTrend detection
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved