Biometrics & Population Modeling IV

Contributed Paper
ROOM: HCCC, Room 20

8:10AM Not Just Dusty Data: What Can We Learn From Range-Wide Analyses of Lesser Prairie-Chicken Morphology?
Carly Aulicky; David A. Haukos; Kent Fricke
Individual populations across a species’ range can vary dramatically in morphology and behavior. Variation in morphological characteristics can indicate differences in local selection pressures such as climate and resource availability. Morphology and body size act as a reflection of resource availability during periods of growth, and directly relate to individual potential energy reserves through mass. For boom and bust species such as the lesser prairie-chicken (Tympanuchus pallidicinctus), morphology provides a useful tool to understand available nutrient resources and environmental differences across the species’ range. The lesser prairie-chicken range is divided among four ecoregions that occur across a dramatic longitudinal precipitation gradient, which subjects local populations to different growing seasons, precipitation levels, and resource availability across the range. We analyzed wing chord, mass, pinnae, tarsus, and tail measurements to assess how and when morphology varies in relationship to precipitation across the species’ range from the mid-1980s to present. We tested separate models by ecoregion, age class, and sex due to significant morphological differences between age and sex classifications. We estimated lipid and fat content and nutritional reserves using ash weight validated equations and morphometrics collected across the lesser prairie-chicken range. Mass explains most morphological variation for males and females for all four ecoregions. Mass, wing, pinnae, tarsus, and tail differed among ecoregions, with no clear trend among ecoregion, age, or sex. Adult morphometrics in the Sand Shinnery Oak Prairie Ecoregion were affected by the preceding year’s annual precipitation, with more adult male morphometrics (mass, wing, pinnae, and tarsus) affected than females (mass). Range-wide analyses of lesser prairie-chicken morphology provide insight to resource allocation during growth in response to precipitation and non-invasive estimations of body condition.
8:30AM Drought Governs Extinction of an Arid-Land Amphibian: Insights From a Spatially Explicit Occupancy Model
Erin R. Zylstra; Don E. Swann; Blake R. Hossack; Erin Muths; Robert J. Steidl
Understanding how natural and anthropogenic processes affect dynamics of species with patchy distributions is necessary to predict how populations will respond to future environmental changes, especially in light of climate change. To facilitate exploring how climate and other factors influence distributional dynamics, we extended a spatially explicit occupancy model to allow extinction and colonization probabilities to vary over space and time as a function of covariates, a framework that has application to many species that inhabit patchy environments. We applied the model to 22 years of biannual survey data for lowland leopard frogs (Lithobates yavapaiensis), an amphibian that inhabits arid mountain canyons in southern Arizona. Local populations were more likely to go extinct (Pr[extinction] > 0.11) during periods of drought, particularly at sites that lacked deep or spring-fed pools that held water reliably. Unoccupied sites were less likely to be colonized (Pr[colonization] ≤ 0.12) when larval or dispersal periods were drier than normal, decreasing the probability of individuals metamorphosing and potential emigrants reaching new sites. Further, colonization rates were low in areas where sediments released by wildfires exceeded 25% of pool volume. Between 1996 and 2008, frogs occupied 20% to 61% of sites in each of the two watersheds we surveyed. Between 2009 and 2017, however, occupancy in one of the watersheds declined markedly, and we observed no frogs in this watershed after 2015. Although a multi-year drought affected dynamics of frogs in both watersheds, frogs persisted only where sediment levels were low and surface water was reliable. If future increases in drought frequency or severity limit the number of inhabitable sites or reduce the permeability of the landscape to dispersing frogs, population persistence is likely to be reduced or to require intensive management.
8:50AM Multi-Scale Drivers of Amphibian Community Diversity Across National Parks in the National Capitol Region
Alexander D. Wright; Elise F. Zipkin; Evan H. Campbell Grant
U.S. National Parks within the National Capitol Region (NCR) are becoming increasingly urbanized and isolated, such that both internal and external threats may negatively affect park resources. The overall health of parks in the NCR is evaluated by monitoring a set of ‘Vital Signs’, which includes amphibians because of their high sensitivity to environmental change. However, determining which drivers (and their spatial scales of relevance) impact species is often complex and difficult to ascertain, complicating management plans and the appropriate scale for conservation activities. The objective of our study was to estimate amphibian species richness within and among parks, and determine the key drivers of species’ trends over time within the region. To do this, we used monitoring data from the U.S. Geological Survey’s Amphibian Research and Monitoring Initiative, which has actively monitored the lentic amphibian community in 9 of the 11 parks since 2005 (over 3000 surveys at ~ 475 wetlands). We developed multi-region, community occupancy models to estimate species- and region-specific occupancy and detection rates for the amphibian community using the monitoring data. We evaluated the effects of variables on species’ occupancy rates at local scales (e.g. wetland area, wetland connectivity, etc.) as well as the impact of landscape-scale drivers (e.g. park size, park isolation, etc.) on park-level richness. Generally, we found a positive effect of park size and a negative effect of isolation on species richness within parks, and positive effects of wetland area and connectivity on species’ occupancy probabilities at individual wetlands, although results varied by species. Our results demonstrate the multi-scaled complexity driving amphibian population trends, and can aid resource managers in evaluating how conservation actions may mitigate threats to amphibians at both wetland- and park-level scales.
9:10AM Accounting for Animal-Specific Variation in Habitat Selection Studies Using Random Effects: Efficient Estimation Using Inla
John R. Fieberg; Stefanie Muff; Johannes Signer
Popular conceptual frameworks used to study habitat selection include resource selection functions (RSFs) and step-selection functions (SSFs) estimated using logistic and conditional logistic regression, respectively. Both frameworks compare environmental covariates associated with locations that animals visit with environmental covariates at a set of locations assumed available to the animal. The main difference between these approaches is that SSFs condition available points on the current location of the animal, resulting in a stratified data set, whereas RSFs pool all location data and use a single set of available points for each animal. Random-coefficient models are relatively uncommon in habitat selection studies, despite their ability to accommodate inter-individual heterogeneity. One potential explanation is that these models are challenging to fit, particularly when allowing for multiple random slopes. Further, straightforward and efficient one-step procedures for fitting SSFs with random coefficients are currently lacking. We reformulate the conditional logistic regression model as a (likelihood-equivalent) Poisson model with stratum-specific intercepts approximated as a random effect with a large, fixed variance. This allows for fast and accurate computation in a Bayesian framework via integrated nested Laplace approximations (INLA). We compare this approach to other commonly applied alternatives, including random intercept-only models and a two-step algorithm for fitting mixed effects models, using a simulation study and by reanalyzing data from Eurasian otters. Our simulation results demonstrate the importance of including random slopes (not just intercepts) when estimating individual- and population-level habitat selection parameters. In addition, we reinforce the need to weight available points when fitting models as has been previously suggested for RSFs. By providing coded examples, we hope to make efficient estimation of RSFs and SSFs with random effects accessible to anyone in the field.
9:30AM Modeling Multi-Scale Drivers of Northern Pintail Demography in the Prairie Pothole Region
Frances E. Buderman; David N. Koons
Separating the effects of change in anthropogenic land use and climate change is of critical importance for understanding temporal changes in species abundance. However, land-use changes and weather patterns operate at vastly different scales of influence, both in time and in space. Land-use changes and climactic conditions have both been identified as possible contributors to decline in the breeding population of northern pintails (Anas acuta) in the Prairie Pothole Region (PPR). Northern pintails often nest in fallow fields, standing stubble, and uncultivated uplands. However, technological advancements have resulted in more efficient harvest and minimal-tillage practices, resulting in less fallow fields than would be available historically. Northern pintails also prefer to breed near small ephemeral wetlands. However, contemporary irrigation practices often consolidates surface water into larger, deeper, less productive wetlands, with the remaining wetlands being susceptible to drought conditions. To disentangle the effects of land-use changes and climactic conditions, we developed a hierarchical process model that separates small-scale, within-year determinants of settling patterns from large-scale demographic drivers of change in abundance. We also explicitly modeled the detection process and the process by which annual precipitation and temperature regimes contribute to the number of available wetlands, allowing us to evaluate the direct and indirect effects of climate on waterfowl. We applied this model to 50 years of the Breeding Waterfowl and Habitat Survey data for northern pintail and pond counts. We determined that winter and spring precipitation had a strong effect on pond numbers and settling dynamics, but the demographic effect on abundance the following year dissipated over time (i.e., became decoupled). Percentage of cultivated crop land had no effect on settling patterns, but had a negative impact on abundance the following year. The strength of these effects also varied regionally, which has strong implications for prioritizing waterfowl conservation decisions.


Contributed Paper
Location: Huntington Convention Center of Cleveland Date: October 11, 2018 Time: 8:10 am - 9:50 am