BMP Meta-analysis

We reviewed 560 papers that we retrieved during the literature search. Among these papers, 151 had quantifiable effect sizes. We were able to trace the effect sizes of an additional 28 papers in Data Thief III. Among the papers that fit our inclusion criteria, 97 papers assessed the impacts of BMPs on abundance, 46 on nest success, and 37 on species richness. There were not enough retrieved papers that addressed the effects of BMPs on survival (n = 4) and occupancy (n = 5) to be included in this analysis.

We hunted (Tara) and gathered (Brian) a whopping 2011 effect sizes. Phew!

Results

Species richness

The data support that plant NWSG, edge and shrub habitat, provide overwintering habitat, and eliminate pesticides are positively associated with species richness. The data did not support the influence of any other BMP on species richness.

Nest success

Across species, delay hay was the only BMP that had a significant positive effect on nest success. The data did not support the influence of any other BMP that could be evaluated on the probability of nest success.

When assessed by species-habitat association (i.e., grassland obligate, facultative grassland, and shrubland species), only delayed hay was significant.

Abundance

The data support that delay hay, eliminate pesticides, install nest boxes, provide overwintering habitat, plant NWSG, and prescribed fire are positively associated with abundance. Among these, delaying the haying date had the greatest impact on abundance. Across species, the data did not support the influence of any other BMP on abundance.

When assessed by species-habitat association, edge and shrub habitat had significantly positive effects on facultative grassland and shrubland species, but significantly negative effects on obligate grassland species. Conversely, prescribed fire and remove non-native species had significantly negative effects on facultative grassland and shrubland species, but significantly positive effects on obligate grassland species. Likewise, plant NWSG had significantly positive effects on facultative grassland and shrubland species, and no effect on obligate species, while provide overwintering habitat had no effect on facultative grassland and shrubland species, and significantly positive effects on obligate grassland species.

The data support that BMP recommendations are species-specific. BMPs can have conflicting effects on a given species. For example, common yellowthroat abundance is positively associated with plant NWSG, but negatively associated with prescribed fire. Additionally, a given BMP can have conflicting effects on multiple species. For example, prescribed fire has a significantly positive effect on bobolink, grasshopper sparrow, and western meadowlark, but a significantly negative effect on common yellowthroat. Similarly, a given BMP can have a positive effect on some species but no effect on others. For example, manage in patches has a significant positive effect on grasshopper sparrow abundance, but no effect on bobolink abundance.

Effect size plots

Pooled effect size plots

Pooled effect sizes by species class

Pooled effect sizes by species

Distributions

For the distribution plots below, the effect size values are not weighted by the sample size of a given study. This means that studies with lower sample sizes may be over-represented whereas those with higher sample sizes may be underrepresented. The pooled effect sizes (above) account for this and are therefore a much better indication of the error. As such, we recommend that these distribution plots be included as supplemental materials within the manuscript.