Reproductive success is central to long-term population persistence, yet it is rarely assessed in ecosystem restoration. We performed a multilevel meta-analysis to assess whether bird breeding success increases from croplands to restored grasslands or from restored grasslands to natural grasslands. We quantified five comparisons of two breeding success variables between croplands and restored grasslands in three species, and 228 comparisons of nine variables between natural and restored grasslands in 22 species using 14 datasets from North America and Europe. For the few grassland birds that nest in croplands, clutch size was higher in restored fields than in croplands. Most importantly, breeding success did not differ between restored and natural grasslands in nine species and for all species combined. In 13 species, breeding success differed between restored and natural grasslands, and it was often higher in restored grasslands during early breeding stages, but higher in natural grasslands during later stages. Eight species did better in restored grasslands (Clay-colored Sparrow, Dickcissel, Eastern Meadowlark, Lesser Prairie-chicken, Mourning Dove, Red-winged Blackbird, Sprague's Pipit, Western Meadowlark), whereas four species did better in natural grasslands (Baird's Sparrow, Common Nighthawk, Lark Bunting, Savannah Sparrow). Our results suggest that bird breeding success can be as high in restored grasslands as in natural ones, and certainly higher than in croplands. While restored grasslands differ widely by geography, climate, method and management, our results offer hope for ecosystem restoration. However, the differences found also imply that species requirements should be considered in grassland ecosystem restoration and management.
