Variation in nutrient allocation can influence the timing of breeding and ultimately reproductive output. Time and space constraints might exist, however, if fewer food resources are available to meet the costs of reproduction early during the reproductive season. Here, for the first time, we test whether nutrient allocation strategies for reproduction in a shrub-dependent avian species differs with timing of breeding in different ecoregions: a high-elevation landscape, containing spatially complex vegetation (Rocky Mountains) versus a low-elevation, more homogenous landscape (Great Plains). We analyzed data collected from radio-telemetry and stable isotopes to assess the degree to which endogenous (body) reserves are used for reproduction and whether variation in allocation strategies was associated with time of year, ecoregion, habitat quality (including sagebrush type and plant greenness), or maternal characteristics. Using a Bayesian statistical framework, we found that females relied on a similar amount of endogenous reserves for reproduction in first nesting and renesting attempts. Additionally, endogenous contributions declined more rapidly throughout the nesting season in the Rocky Mountains than in the Great Plains. Individuals in high- and intermediate-elevation sagebrush types in the Rocky Mountains used similar amounts of endogenous reserves, whereas females nesting in low-elevation sagebrush used less. Females nesting at intermediate elevations, which experience the greatest flush of new green vegetation during the nesting season, switched their reliance from endogenous to exogenous sources for reproduction as green vegetation became available during spring. Our study highlights adaptations of a nutrient-allocation strategy across areas with varying levels of resources in time and space in a habitat specialist bird. Nutrient allocation by individuals residing in high-elevation areas favors a strategy that mainly uses nutrients gained from wintering habitats, whereas individuals residing in low-elevation areas mainly use exogenous sources for reproduction.

Data were collected across two sagebrush-dominated ecoregions: The Great Plains of eastern Montana, and the Rocky Mountains of southwest Montana, USA (Figure 1). The Great Plains site is located near the town of Roundup, which contains rolling topography ranging in elevation from 975m to 1250m. The dominant sagebrush species are Wyoming big sagebrush (Artemisia tridentata ssp. wyomingensis) and silver sagebrush (A. cana ssp. cana), with an understory of native perennial bunch grasses and forbs. Average annual precipitation and temperature is 359 mm and 9.2^oC, respectively (National Climatic Data Center 2019). The Rocky Mountain site is located near the town of Lakeview in southwest Montana. Sagebrush in this ecoregion occurs across an elevation gradient of 2,000 m to 2,900 m. Mountain big sagebrush (A. tridentata ssp. vaseyana; hereafter; MTSA) occurs at the highest elevations where snowmelt and vegetation green-up occur latest during the spring. Three-tip sagebrush (A. tripartita; hereafter; TTSA) occurs on south-facing mid-elevation slopes where snowmelt and vegetation green-up occur earliest. Basin big sagebrush (A. tridentata ssp. tridentata; hereafter; BBSA) occurs in valley bottoms with deep soils where drifting of snow results in a later snowmelt date, but the earliest vegetation senescence. Average annual precipitation and temperature at Lakeview is 500mm and 1.6^oC, respectively.

We captured female sage-grouse using dipnets and spotlights. Upon capture, we collected blood samples from individual female sage-grouse by brachial venipuncture using a 23-gauge needle with 3 mL syringe. We collected blood samples from a subsample of individual female sage-grouse from the Great Plains site in both 2016 and 2017 (n=10 females each year), and from every captured individual at the Rocky Mountain site from 2014 to 2019 (n=91). We centrifuged blood samples within 12 hours of capture and separated the plasma and red blood cells constituents into separate vials. Samples were stored frozen until we prepped samples for stable isotope analysis.

On each study site, we tracked individual female grouse 2–3 times per week during the pre-nesting season (April to June) until we discovered a nest. Once we discovered a nest, we monitored nest status two times per week until the nest either hatched or failed. We considered a nest to be successful if ≥1 egg hatched as determined by the presence of any of the following in the nest bowl: a chick, an intact egg membrane, and/or eggshell cap. We counted membranes after a nest hatched and estimated when the first egg was laid (hereafter; initiation date) by assuming an egg laying interval of 1.5 days and a 28-day incubation period to determine the date (Moynahan et al. 2007).  For failed nests, we continued to track the female until mid-June to determine whether a second nest was initiated. We included only individuals who successfully hatched young, since our primary focus was on explaining variation in nutrient allocation to offspring production.

If a nest was successful, we captured chicks when they were 2 to 8 days old (average = 3 d) and collected down feathers for stable isotope analysis. Downy feathers from the breast are present at the time of hatch and are predominantly keratin that reflect nutrient (protein) pools used by adult females during egg development (Williams 2012). We captured chicks by hand in the early morning by flushing the brooding female off her dependent young. We stored feather samples from individual chicks in separate paper envelopes prior to stable isotope analysis.