Joshua H. Schmidt

A General Bayesian Hierarchical Model for Estimating Survival of Nests and Young

ABSTRACT. Models for estimating survival probability of nests and young have changed dramatically since the development of the Mayfield method. Improvements in software and a steady increase in computing power have allowed more complexity and realism in these models, allowing researchers to provide better estimates of survival and to relate survival rates to relevant covariates. However, many current analysis methods utilize fixed-effects models with the implicit assumption that the covariates explain all of the variation in the data, other than random variation within a specified family of distributions. This is generally a strong assumption, and, in the presence of heterogeneity and lack of independence, these estimates have been shown to be negatively biased. Others have begun to explore random-effects models for these situations, but a readily applicable Bayesian approach has been lacking. We present a general Bayesian modeling framework appropriate for survival of both nests and young that simultaneously allows for the inclusion of individual covariates and random effects and provides a measure of goodness-of-fit. We used previously published data on survival of Common Goldeneye (Bucephala clangula) ducklings in interior Alaska and on nest survival in three species of prairie-nesting ducks that nested in the Missouri Coteau region of North Dakota to demonstrate this approach. The inclusion of a brood-level random effect in the Common Goldeneye example increased point estimates and credible interval [CI] coverage from 0.62 (95% CI: 0.49–0.73) and 0.66 (95% CI: 0.58–0.74) for 2002 and 2003, respectively, to 0.69 (95% CI: 0.42–0.88) and 0.74 (95% CI: 0.57–0.88) for 2002 and 2003, respectively.

Landscape-scale patterns in tree occupancy and abundance in subarctic Alaska

Recent studies suggest that climate warming in interior Alaska may result in major shifts from spruce-dominated forests to broadleaf-dominated forests or even grasslands. To quantify patterns in tree distribution and abundance and to investigate the potential for changes in forest dynamics through time, we initiated a spatially extensive vegetation monitoring program covering 1.28 million ha in Denali National Park and Preserve (DNPP). Using a probabilistic sampling design, we collected field measurements throughout the study area to develop spatially explicit Bayesian hierarchical models of tree occupancy and abundance. These models demonstrated a strong partitioning of the landscape among the six tree species in DNPP, and allowed us to account for and examine residual spatial autocorrelation in our data. Tree distributions were governed by two primary ecological gradients: (1) the gradient from low elevation, poorly drained, permafrost-influenced sites with shallow active layers and low soil pH (dominat...

Estimating demographic parameters using a combination of known‐fate and open N‐mixture models

Accurate estimates of demographic parameters are required to infer appropriate ecological relationships and inform management actions. Known-fate data from marked individuals are commonly used to estimate survival rates, whereas N-mixture models use count data from unmarked individuals to estimate multiple demographic parameters. However, a joint approach combining the strengths of both analytical tools has not been developed. Here we develop an integrated model combining known-fate and open N-mixture models, allowing the estimation of detection probability, recruitment, and the joint estimation of survival. We demonstrate our approach through both simulations and an applied example using four years of known-fate and pack count data for wolves (Canis lupus). Simulation results indicated that the integrated model reliably recovered parameters with no evidence of bias, and survival estimates were more precise under the joint model. Results from the applied example indicated that the marked sample of wolves was biased toward individuals with higher apparent survival rates than the unmarked pack mates, suggesting that joint estimates may be more representative of the overall population. Our integrated model is a practical approach for reducing bias while increasing precision and the amount of information gained from mark-resight data sets. We provide implementations in both the BUGS language and an R package.

Scale dependence in occupancy models: implications for estimating bear den distribution and abundance

Monitoring programs are typically designed to identify long-term trends in animal abundance, however estimating abundance at a relevant scale can be logistically prohibitive. This is particularly true for species that occur at low densities or those with large home ranges. In such cases, occupancy surveys are often employed in place of more expensive abundance estimation techniques such as mark-recapture because precise estimation of occupancy probability generally requires fewer data. Although choice of plot size is a critical design element of occupancy monitoring, relatively little effort has been expended to develop or test plot size recommendations. Animal movement between surveys can complicate efforts to obtain an optimal plot size, but surveys of fixed objects, such as nests, dens, or burrows can provide insight about scale effects because the population exposed to sampling does not change during the duration of the survey. We used repeated aerial occupancy surveys to obtain estimates of brown bear (Ursus arctos) den distribution and abundance in a portion of Katmai National Park and Preserve in Alaska. We then used these data to assess the importance of plot size selection and highlight the effects of spatial grain on the resulting inference and utility for monitoring. Scale effects in estimates of mean den-based site occupancy, but not total den abundance demonstrated that careful selection of sample unit size is important if estimating occupancy probability is a primary monitoring objective. We expect occupancy surveys based on important structures such as nests or dens could have wide applicability for many species.

Habitat use and population status of Yellow-billed and Pacific loons in western Alaska, USA

ABSTRACT Effective conservation of sympatric avian populations depends on unbiased estimates of population size, distribution, and habitat use. For populations of Yellow-billed Loons (Gavia adamsii) and Pacific Loons (G. pacifica) co-occurring in Arctic wetland communities in Alaska, USA, such data are limited and difficult to obtain, hindering population assessments and decision making. The Yellow-billed Loon is also under consideration for additional protections under the Endangered Species Act due to small global population size, specific habitat requirements, and low fecundity, further increasing the need for information at the landscape scale. To help evaluate the population status and habitat use of both species, we used repeated aerial surveys and a dynamic multistate occupancy modeling approach to jointly estimate 1) probability of lake use and 2) probability of use for nesting for Yellow-billed and Pacific loon populations at the landscape scale on the Seward Peninsula and Cape Krusenstern, Alaska, in 2011 and 2013. We also estimated state-specific transition probabilities and degree of interspecific competition to assess population stability and degree of species interactions. We found that probability of site reuse (φYellow-billed = 0.73 [0.44–0.94]; φPacific = 0.86 [0.72–0.98]) or reuse for nesting (φYellow-billed = 0.72 [0.46–0.97]; φPacific = 0.59 [0.38–0.85]) in 2013 was high, as was overall use of lakes >7 ha by loons (>80%). These results suggested that lake habitats may have been saturated, and that populations of both species were stable over the two-year interval between surveys. Our estimates indicated that nesting populations in western Alaska were much larger than previously thought for both Yellow-billed (∼2.5 times larger) and Pacific loons (∼1.5–2.0 times larger). Together our results indicate that Arctic wetlands in western Alaska are important for both species and that loon populations in this area warrant additional consideration for conservation.

Possible Secondary Population-Level Effects of Selective Harvest of Adult Male Muskoxen

Selective harvest regimes are often focused on males resulting in skewed sex-ratios, and for many ungulate species this strategy is sustainable. However, muskoxen (Ovibos moschatus) are very social and mature bulls (≥4 years old), particularly prime-age bulls (6–10 years old), play important roles in predator defense and recruitment. A year-round social structure incorporating large males into mixed-sex groups could make this species more susceptible to the effects of selective harvest if population composition and sex-ratios influence overall survival and reproductive success. Using detailed data collected on the muskox population occupying the Seward Peninsula, Alaska during 2002–2012, we formulated the hypothesis that the selective harvest of mature bulls may be related to documented changes in population composition and growth rates in this species. In addition, we reviewed existing published information from two other populations in Alaska, the Cape Thompson and Northeastern populations, to compare population growth rates among the three areas under differential harvest rates relative to our hypothesis. We found that on the Seward Peninsula, mature bull:adult cow ratios declined 4–12%/year and short-yearling:adult cow ratios (i.e., recruitment) declined 8–9%/year in the most heavily harvested areas. Growth rates in all 3 populations decreased disproportionately after increases in the number of bulls harvested, and calf:cow ratios declined in the Northeastern population as harvest increased. While lack of appropriate data prevented us from excluding other potential causes such as density dependent effects and changes in predator densities, our results did align with our hypothesis, suggesting that in the interest of conservation, harvest of mature males should be restricted until causal factors can be more definitively identified. If confirmed by additional research, our findings would have important implications for harvest management and conservation of muskoxen and other ungulate species with similar life-histories.

Season length influences breeding range dynamics of trumpeter swans Cygnus buccinator

Abstract The breeding range of large-bodied waterfowl nesting in the northern boreal forest is likely influenced by breeding season length. This may be particularly true for the largest species of North American waterfowl, the trumpeter swan Cygnus buccinator, due to the extended time period necessary to raise young to fledging. This species recently recovered from near-extinction in the early 1900s to reoccupy historic breeding areas throughout the boreal forest in Alaska, although recolonization patterns may have been influenced by variation in season length over the same time period. This may have resulted in range expansion into areas that were historically unavailable due to an ice-free period insufficient for successful reproduction. We used hierarchical occupancy models to analyze trumpeter swan survey data collected over the entire breeding range in Alaska during 1968-2005. We fit models containing combinations of recolonization parameters, trend and latitude, and season length to these data to determine whether these variables explained the variation in occupancy across our survey area. Support for season length parameters would provide evidence that the recolonization process was partially related to the length of the breeding season. We expected that occupancy probability would increase range-wide due to overall population growth, while occupancy would be greatest at mid-latitudes, near the center of the species range. Because this population was recovering, we also expected that expansion would proceed outward from the range center. Our results indicated that habitat occupancy was positively related to season length, partially explaining the recently observed northward range expansion. Our results suggest that increases in annual temperatures due to climate warming would likely be associated with further range expansion in trumpeter swans and may have implications for other wetland obligates. Changes in species distributions will likely increase competition for breeding areas with potential negative effects on species not limited by season length. This may already be occurring in Alaska where the breeding distribution of trumpeter swans has begun to overlap with that of tundra swans Cygnus columbianus.

Weather-driven change in primary productivity explains variation in the amplitude of two herbivore population cycles in a boreal system

Vertebrate populations throughout the circumpolar north often exhibit cyclic dynamics, and predation is generally considered to be a primary driver of these cycles in a variety of herbivore species. However, weather and climate play a role in entraining cycles over broad landscapes and may alter cyclic dynamics, although the mechanism by which these processes operate is uncertain. Experimental and observational work has suggested that weather influences primary productivity over multi-year time periods, suggesting a pathway through which weather and climate may influence cyclic herbivore dynamics. Using long-term monitoring data, we investigated the relationships among multi-year weather conditions, measures of primary productivity, and the abundance of two cyclic herbivore species: snowshoe hare and northern red-backed vole. We found that precipitation (rain and snow) and growing season temperatures were strongly associated with variation in primary productivity over multi-year time horizons. In turn, fourfold variation in the amplitude of both the hare and vole cycles observed in our study area corresponded to long-term changes in primary productivity. The congruence of our results for these two species suggests a general mechanism by which weather and climate might influence cyclic herbivore population dynamics. Our findings also suggested that the association between climate warming and the disappearance of cycles might be initiated by changes in primary productivity. This work provides an explanation for observed influences of weather and climate on primary productivity and population cycles and will help our collective understanding of how future climate warming may influence these ecological phenomena in the future.

Bottom-up processes drive reproductive success in an apex predator

Abstract One of the central goals of the field of population ecology is to identify the drivers of population dynamics, particularly in the context of predator–prey relationships. Understanding the relative role of top‐down versus bottom‐up drivers is of particular interest in understanding ecosystem dynamics. Our goal was to explore predator–prey relationships in a boreal ecosystem in interior Alaska through the use of multispecies long‐term monitoring data. We used 29 years of field data and a dynamic multistate site occupancy modeling approach to explore the trophic relationships between an apex predator, the golden eagle, and cyclic populations of the two primary prey species available to eagles early in the breeding season, snowshoe hare and willow ptarmigan. We found that golden eagle reproductive success was reliant on prey numbers, but also responded prior to changes in the phase of the snowshoe hare population cycle and failed to respond to variation in hare cycle amplitude. There was no lagged response to ptarmigan populations, and ptarmigan populations recovered quickly from the low phase. Together, these results suggested that eagle reproduction is largely driven by bottom‐up processes, with little evidence of top‐down control of either ptarmigan or hare populations. Although the relationship between golden eagle reproductive success and prey abundance had been previously established, here we established prey populations are likely driving eagle dynamics through bottom‐up processes. The key to this insight was our focus on golden eagle reproductive parameters rather than overall abundance. Although our inference is limited to the golden eagle–hare–ptarmigan relationships we studied, our results suggest caution in interpreting predator–prey abundance patterns among other species as strong evidence for top‐down control.

Improving inference for aerial surveys of bears: The importance of assumptions and the cost of unnecessary complexity

Abstract Obtaining useful estimates of wildlife abundance or density requires thoughtful attention to potential sources of bias and precision, and it is widely understood that addressing incomplete detection is critical to appropriate inference. When the underlying assumptions of sampling approaches are violated, both increased bias and reduced precision of the population estimator may result. Bear (Ursus spp.) populations can be difficult to sample and are often monitored using mark‐recapture distance sampling (MRDS) methods, although obtaining adequate sample sizes can be cost prohibitive. With the goal of improving inference, we examined the underlying methodological assumptions and estimator efficiency of three datasets collected under an MRDS protocol designed specifically for bears. We analyzed these data using MRDS, conventional distance sampling (CDS), and open‐distance sampling approaches to evaluate the apparent bias‐precision tradeoff relative to the assumptions inherent under each approach. We also evaluated the incorporation of informative priors on detection parameters within a Bayesian context. We found that the CDS estimator had low apparent bias and was more efficient than the more complex MRDS estimator. When combined with informative priors on the detection process, precision was increased by >50% compared to the MRDS approach with little apparent bias. In addition, open‐distance sampling models revealed a serious violation of the assumption that all bears were available to be sampled. Inference is directly related to the underlying assumptions of the survey design and the analytical tools employed. We show that for aerial surveys of bears, avoidance of unnecessary model complexity, use of prior information, and the application of open population models can be used to greatly improve estimator performance and simplify field protocols. Although we focused on distance sampling‐based aerial surveys for bears, the general concepts we addressed apply to a variety of wildlife survey contexts.