|Abstract or Summary
- In grasslands, grazing by large ungulates can influence vegetation structure, composition, primary productivity, and ultimately, ecosystem functioning. While grazing represents a complex disturbance, grazing intensity largely determines the effects of
grazing on vegetation. Structural and compositional changes in the plant community caused by grazing could have bottom-up effects on species and interactions at higher trophic levels. Thus, particular management strategies for domestic livestock in rangeland systems could exert a strong affect on grassland wildlife. Grassland-dependent songbirds may be particularly susceptible to the effects of domestic grazers because they depend on grassland vegetation for foraging and nesting. Domestic livestock may influence grassland-breeding bird populations by affecting settlement decisions, resource
availability, or reproductive success. We investigated the effects of grazing intensity on grassland vegetation structure and songbird demography in a northwestern bunchgrass prairie using paddocks with experimentally-manipulated cattle stocking rates. We compared effects of four stocking rates (0, 14.4, 28.8, and 43.2 animal unit months) on songbirds using a randomized complete block design with four replicates of each stocking rate to address hypotheses regarding demography of grassland songbirds. Overall paddock-level vegetation structure decreased and structural heterogeneity
of vegetation increased with higher stocking rates, and those effects carried over one-year post-grazing. However, most bird species were able to locate nesting sites with similar vegetation structure regardless of paddock-level effects of stocking rate. The exceptions were western meadowlarks and vesper sparrows; nests of these species in paddocks with higher stocking rates had less vegetative cover. Apparent nest density for grasshopper sparrows was negatively affected by higher stocking rates. Grazing treatment effects on songbird population density were restricted to negative effects of higher stocking rates on savannah sparrows, but this relationship was not observed until the post-treatment year.
Songbird community composition differed between control and heavily-grazed paddocks, but diversity was not affected by stocking rate.
Nest fates were evaluated to determine whether stocking rate influenced nest survival or cause-specific nest failure. Other variables such as vegetation structure and predator community, date, year, and nest age were included to help clarify which mechanisms might be responsible for differences in nest survival or failure rates among treatments. For our analysis, we introduce the use of a novel software package, McNestimate, to estimate the daily probability of nest survival and failure from specific causes. McNestimate estimates the probability of nest failure from competing causes when the exact dates of failure are unknown using a Markov Chain framework, and incorporates a model selection approach which allows the use of covariates to help identify variables important in explaining variation in the daily probability of nest failure. Nest predation rates increased with the age of the nest and throughout the breeding season, but were not affected by stocking rate. The probability of nest failure from adverse weather declined throughout the season, but the rate of decline depended on year. Nest failure rates due to trampling were higher in paddocks with higher stocking rates, but also depended on the number of days cattle were present during the nesting period. Patterns of overall probability of nest success were driven by predation patterns in the first year, but in the second year were strongly influenced by the chances of weather-related nest failure. Although starvation was not identified as a significant source of nest failure, grazing-induced changes to vegetation structure and composition could influence food availability for breeding songbirds, ultimately affecting the composition of nestling diets and nestling condition. To better understand the relationship between grazing intensity, nestling diet composition, and subsequent effects on nestling condition, we examined the invertebrate composition of nestling fecal samples. All species showed strong preferences for Lepidoptera (moths and butterflies) larvae, and partial preferences for Coleoptera (beetles) and Araneae (spiders). The proportion of preferred prey items was not affected by stocking rate. There were effects of bird species on the proportion of Araneae and Coleoptera and the proportion of Acrididae (short-horned grasshoppers) in the diet of western meadowlark nestlings decreased with high stocking rates. Growth rates for western meadowlarks and vesper sparrows were negatively affected by higher stocking rates. These results suggest that stocking rates can have variable effects on grassland songbird population and nest density depending on each species' habitat
requirements. However, negative effects of high stocking rates on nest survival and nestling condition could have consequences for juvenile survival and recruitment. Overall, low-to-moderate stocking rates are likely compatible with many grassland bird species in northwest bunchgrass prairie, and although heavier livestock grazing may help create suitable vegetation structure for some songbird species, high stocking rates may influence grassland songbird diet quality, or have negative effects on nestling condition. We hypothesized that grazing intensity could influence the grassland songbird community through "bottom-up" effects on vegetation, but effects of grazing at different intensities did not translate directly through the food web to influence songbird populations as strongly as lower trophic levels. Processes responsible for changes in community composition such as immigration or emigration may not have had time to ensue during our short-term experiment; alternatively, sufficient spatial or temporal heterogeneity remained in the system, even at the highest grazing intensity, such that
grazing-induced changes in lower trophic levels were irrelevant for most songbird species. Our results contribute to understanding grassland songbird demographic responses to different grazing intensities and identify specific mechanisms by which conservation measures for declining grassland bird populations can be improved.