Megafires in the 2019-20 fire area: a proposal to collect site-level data from burned sites in the megafire region 62
We refer to megafires as those more than 100,000 ha, with 12 megafires recognized throughout the 2019–20 fire area5, although proposed terminology would classify these as giga- and tera-fires61. We invited contributors to join our collaboration and gathered site-level data for 62 projects. The Wildlife and Habitat Bushfire Recovery Program provided funding for some of the projects that we email leaders of. We promoted our proposal on social media in summer 2021–2022, ran a workshop at the Ecological Society of Australia conference in November 2021 and held a 2-day online symposium in February 2022. After potential contributors contacted us, projects were compiled throughout the year. Projects had to include site-level data from sites burnt in 2019–2020 within the megafire region62. To enable means and s.d., a minimum of three burned and unburnt sites or at least three non-zero sites were needed. Projects also needed to report either an estimate or index of abundance (44 projects), occurrence (17 projects) or probability of occurrence after accounting for detection (1 project). Although we canvassed researchers for a full range of response variables, the datasets provided included almost exclusively presence–absence or abundance data. Both Abundance and occurrence effect sizes were justified together. Only one project met all other criteria but reported responses that were not related to abundance or occurrence (effects of grazing on reproduction in one orchid species), and thus was omitted from analysis. Five other projects were excluded because they did not have at least three unburnt sites.
The findings and conclusions in this essay are those of the author and should not be construed to represent any official USDA or US Government determination or policy.
Assessing biodiversity impacts of the 2019 -2020 Australian megafires with meta-analysis models: The mean change and difference effects did not change after the fires
There were only small changes in mean effect sizes when we fitted the meta-analytical models to evaluate the year effect. 2b), suggesting positive and negative impacts of the fires had not substantively changed 2 years after the fires. When there were several survey times after the fire, we averaged responses across time periods for sites within projects.
Some variances for mean change and mean difference effects were zero. The meta-analysis models weren’t allowed to converge. To enable models to converge and use a dataset similar to the SMCC and SMD, we removed effects with zero variance before analysis. = 2.3%; Supplementary Table 1).
The differences in effect sizes were evaluated using the first and second period, year 1 and year 2. We used time gaps between study periods to define surveys that took between 1.5 and 2.5 years to measure response after fire, and surveys that took between 1.5 and 2.5 years to examine responses. The year 1 and year 2 effects were only partially overlap because we used the timegaps in the survey dates to maximize the data that could be included for each project. Ten datasets had a long period of data excluded.
Before using the MC and MD functions data for species within a project was standardized, so that each effect size was calculated from zero to one. While this makes data with different ranges of values comparable, it also risks biasing down effect sizes when there are large outliers. We therefore also calculate MC and MD after removing outliers. We assumed our count data (x) approximate a Poisson distribution, and by transforming our data as y = 2√x, outliers can be identified as those greater than (median(y) + 3), based on the approximation that y is normally distributed with a s.d. of 1 (ref. 69). This approach is conservative, such that if the Poisson lambda value is less than three, it will identify fewer outliers. It doesn’t identify any outliers when applied to binomial data. When outliers were included, results were similar, but effect sizes were slightly larger.
Source: Biodiversity impacts of the 2019–2020 Australian megafires
The After-Control-Impact Three-Dimensional Design for the Multi-Vector Detector at Fermilab
The third design was part of the after–control-impact projects. The control and impact designs were converted to burnt by subtracting the before fire response from the after fire response. Subsequently, the SMD was calculated using the same method as the second design.