## Optimized Frequency Measures for Monitoring Trends in Tallgrass Prairie > [!Cite]- > DeBacker, Michael D., John S. Heywood, and Lloyd W. Morrison. “Optimized Frequency Measures for Monitoring Trends in Tallgrass Prairie.” _Rangeland Ecology & Management_ 64, no. 3 (May 2011): 301–8. [https://doi.org/10.2111/REM-D-09-00179.1](https://doi.org/10.2111/REM-D-09-00179.1). ## Notes %% begin notes %% %% end notes %% %% begin annotations %% ## Annotations | evernote 2016.05.05 Estimates of population abundance for most tallgrass prairie species are not informative given their scarcity across space and time. For managers of protected lands, the relevant information is whether these species persist in the community over time. Multiscale sample designs are effective for monitoring species persistence by capturing uncommon species in large plots. At the same time, multiscale designs provide useful data for monitoring the abundance of common species through their frequency of occurrence in smaller plots (Peet et al. 1998; Stohlgren et al. 1998; James et al. 2009). we have shown that frequency data are most sensitive to changes in underlying plant density when they are acquired using a plot size that yields an average frequency between 20% and 50% (Heywood and DeBacker 2007). we investigate 1) decision rules for assigning a plot frame size that yields an optimized frequency measure for abundant species, 2) the reliability of an assigned frame size to deliver species frequency measures within an acceptable range over time, and 3) the aggregation of results from multiple spatial scales and species into useful information products for land managers. The frequency (percentage of plots occupied per sample site) of a particular species depends on the spatial scale over which it is sampled (i.e., plot size), and is sensitive to changes in underlying density. The optimal plot size was determined for each of 27 abundant species (i.e., all species occurring at .15% frequency at the 10- m 2 scale) by assigning species to a plot frame size that yielded, on average over the 5-yr study, a frequency nearest 35% (i.e., the midpoint of the 20–50% target range). Once a plot frame size was assigned, frequency was calculated fromthe same plot frame size each year of the study. The percentage of plots in which a species occurred at this spatial scale was defined as the species’ optimized frequency for that year. We determined the percentage of each species’ annual optimized frequency measures falling within 620% of the 35% target (i.e., 15–55%), which was defined as an acceptable range of frequencies. An arc-sine square-root transformation was applied to normalize the proportional data (Kutner et al. 2005). From these data, we calculated the trajectory of yearly change, defined as the average across species of the change in optimized frequency between pairs of consecutive years; and the amplitude of yearly change, defined as the average across species of the absolute value of the change in optimized frequency between consecutive pairs of years. These tools allow land managers to compare trends among species of interest in such a way that changes in frequency represent an equivalent, proportionate change in the species’ abundance (e.g., Fig. 3). That is, those species with smaller optimal plot sizes are more abundant than those species with larger optimal plot sizes; however, comparable magnitudes of change from year to year for different species represent an equal proportion of each species’ population (which would amount to more individuals of very abundant species). No specific combination of trajectory (i.e., bigger populations vs. smaller populations) and amplitude (broader vs. narrower ranges) is arguably preferable. However, managers may be most interested in extreme values that may indicate strong directional changes in population size and/or a large magnitude of change from year to year, either of which may indicate a system in transition or going out of control and unable to recover. %% end annotations %% %% Import Date: 2024-08-15T07:50:05.215-06:00 %%