Land‐use and land‐cover change associated with agriculture is one of the main drivers of biodiversity loss. In heavily modified agricultural landscapes, grazing lands may be the only areas that can provide essential resources for native grassland species. Management decisions, such as choice of livestock species, affect the extent to which grazing lands provide suitable habitat for native species such as pollinators. Our study compared how sheep versus cattle herbivory affected floral resources and butterfly abundance across low‐diversity, former Conservation Reserve Program (CRP) pastures managed with patch‐burn grazing. Across all years (2017–2019), flowering species richness and abundance were significantly higher in cattle pastures than sheep pastures. On average, we recorded 6.9 flowering species/transect in cattle pastures and 3.8 flowering species/transect in sheep pastures. The average floral abundance per transect was 1278 stems/transect in cattle pastures and 116 stems/transect in pastures grazed by sheep. Similarly, we observed higher butterfly species richness, diversity, and abundance in cattle than in sheep pastures. In cattle pastures, we observed an average of 75 butterflies and 6.75 species per transect, compared with an average of 52 butterflies and 3.37 species per transect in sheep pastures. However, the butterfly community composition did not significantly differ between grazing treatments likely because agricultural‐tolerant, habitat generalists comprised the majority of the butterfly community. Five generalist butterflies comprised 92.3% of observations;Colias philodicewas the most abundant (61% of observations).Speyeria idaliaandDanaus plexippus, two butterflies of conservation concern, comprised less than 0.5% of butterfly observations. Our results, which are among the first attempt quantifying butterfly use of post‐CRP fields grazed by livestock, show that increased precipitation and cattle grazing promoted higher forb abundance and richness. However, additional interventions may be needed to enhance floral resources to sustain and improve pollinator diversity in these landscapes.

Non-native plants can reduce grassland biodiversity, degrade wildlife habitat, and threaten rural livelihoods. Management can be costly, and the successful eradication of undesirable species does not guarantee the restoration of ecosystem service delivery. An alternative to the eradication of invasive species in rangelands is to target the restoration of diversity and heterogeneous plant structure, which have direct links to ecosystem function. In this study, we evaluate patch-burn grazing (PBG) with one and two fires per year and variably stocked rotational grazing in Poa pratensis- and Bromus inermis-invaded grasslands using traditional (cover) and process-based (diversity and vegetation structural heterogeneity) frameworks in central North Dakota, USA. Within 3–4 years of initiating management, we found little evidence of decreased Poa pratensis and Bromus inermis cover compared to continuous grazing (Poa pratensis F_3,12 = 0.662, p = 0.59; Bromus inermis F_3,12 = 0.13, p = 0.13). However, beta diversity increased over time in all treatments compared to continuous grazing (t_PBG1 = 2.71, t_PBG2 = 3.45, t_Rotational = 3.72), and variably stocked rotational treatments had greater increases in spatial heterogeneity in litter depth and vegetation structure than continuously grazed pastures (t_{visual obstruction}= 2.42, p = 0.03; t_{litter depth} = 2.59, p = 0.02) over the same time period. Alternative frameworks that promote grassland diversity and heterogeneity support the restoration of ecological services and processes in invaded grasslands.

Within agricultural landscapes, native bees often rely on limited natural and seminatural lands to provide the majority of the food and nesting resources that sustain them. To understand better how management can affect pollinators in these seminatural areas, we compared how sheep or cattle herbivory influenced floral resources and bee communities in low‐diversity, former Conservation Reserve Program (CRP) pastures managed with patch‐burn grazing. We sampled bee communities and floral resources three times per season in 2017, 2018, and 2019. We used plant‐pollinator line transect sampling and collected bees and counted all flowering stems within 1 m. Across all years, we found that floral abundance, floral richness, floral diversity (Simpson’s) and bee richness and abundance were significantly higher in cattle pastures compared to sheep. In cattle pastures, 46 native bee species plus honey bees interacted with 25 of 68 available flowering forbs. In sheep pastures, we recorded 14 native bee species and honey bees interacted with 10 of 34 flowering species. Native bee abundance and native bee richness were best explained by models that included an interaction of floral richness and year. Overall, our results suggest that season‐long sheep grazing in low‐diversity grasslands greatly reduces available floral resources and correlates with much lower bee abundance and native bee diversity. Given the importance of pollinators to natural and agricultural systems, it is imperative that we take proactive actions to increase forb richness and native flower abundance in seminatural lands to maintain a more diverse and resilient bee community that can continue to support pollination services and global food security.